* Scan files git calls binary for secrets before push and in pregate (GST-2195)
A NUL byte, UTF-16 text or a -diff attribute made git print 'Binary files differ', so a token in that file passed pregate's secrets check and the pre-push hook. Both diffs now use --text and read such files through scannableText, keeping only credential findings; pregate's secrets check also fails closed when its merge-base diff does not finish.
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Find tokens in UTF-16 text that is mostly Cyrillic or CJK, and note the size limit for binaries
Review-Finding: p1/red-team#1 reader=red-team
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Read each line of a binary file both plainly and as UTF-16, so a nearby byte cannot hide a token
Review-Finding: p2/specialist-batched#1 reader=specialist-batched
Review-Finding: p2/red-team#1 reader=red-team
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep tokens beside CJK characters in UTF-16 files whole, and parse large added diffs in linear time
Review-Finding: p2v/outside-voice#1 reader=outside-voice
Review-Finding: p2v/targeted#1 reader=targeted
Review-Finding: p2v/outside-voice#2 reader=outside-voice
Review-Finding: p2v/adversarial-native#1 reader=adversarial-native
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Split two long sentences in the changelog entry for binary-file scanning
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Describe how the secret check reads binary files, its size limits and its new pass line
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Say which UTF-16 layouts the binary-file secret scan does not catch yet
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Stop the merge queue from calling a head green while a check on it failed (GST-2167)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Don't count a workflow run cancelled before any job as a failed check, and keep a jobless failure counted until its replacement starts (GST-2167)
A concurrency group cancels a pending run when a newer one at another SHA arrives, so the main backstop at 526f604f91 and 5fdfa14904 would have read red forever. A replacing run that is still queued with no jobs no longer hides the failure before it. Adds a test that an unreadable workflow-runs list is unknown.
Review-Finding: p1/red-team#1 reader=red-team
Review-Finding: p1/red-team#2 reader=red-team
Review-Finding: p1/specialist-batched#1 reader=testing
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Judge each check suite by its latest attempt that ran jobs, so a rerun clears a dropped matrix failure and an empty failed rerun stays red (GST-2167)
ciAtHead now reads check runs with the default filter and with filter=all. A suite the default list names is judged by that list alone; a suite only filter=all names is judged by its earlier attempts, and its workflow run counts as a failed check with no jobs when it failed and nothing in the suite already counts the failure.
Review-Finding: v1/outside-voice#1 reader=outside-voice
Review-Finding: v1/outside-voice#2 reader=outside-voice
Review-Finding: v1/adversarial-native#1 reader=adversarial-native
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Name the changelog fragment after the branch and split a run-on sentence in it (GST-2167)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Describe how the merge queue now counts CI in the ship policy doc (GST-2167)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Read the merge queue's notes in one git process so land and ready stay fast (GST-2198)
Rebuilding local queue state from refs/notes/gstack-trains ran one git cat-file per note: 14 to 15 s per land and first ready on the live 1,442 notes, growing about 8 s a week toward the 120 s command timeout. One git cat-file --batch now reads them all (about 120 ms on the same notes) and restores byte-identical records.
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Refuse a failed batched notes read instead of treating it as an empty queue (GST-2198)
When the one git cat-file --batch process fails, times out or overruns its buffer, rebuildFromNotes now returns fetched:false, so land --repair and the freeze lock refuse instead of reading no notes. The test's counting git shim now execs the git in git --exec-path, so a PATH git wrapper cannot loop back into it.
Review-Finding: p1/red-team#1 reader=red-team
Review-Finding: p1/red-team#2 reader=red-team
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Say in the release note that a failed notes read is refused (GST-2198)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Test that a Captain cannot archive a thread still holding open work (GST-2188)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Refuse to archive a thread that holds an open PR and close a landed PR's issue (GST-2188)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Clear the archive check's handoff file on a failed read and say plainly that long PR lists cannot be checked
Review-Finding: p1/specialist-batched#2 reader=testing
Review-Finding: p1/specialist-batched#3 reader=state-machines
Review-Finding: p1/red-team#1 reader=red-team
Review-Finding: p1/red-team#3 reader=red-team
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Refuse to archive on an issue list with no rows, and match tracker codes in any case when closing a landed PR's issues
Review-Finding: p3/outside-voice#1 reader=outside-voice
Review-Finding: p3/outside-voice#2 reader=outside-voice
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Say in the skills guide that a rotated thread is archived only once none of its PRs is open
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Shorten the runaway-stopper archive step so the skill fits its context budget after the rebase
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep Free Tests planning under the job cap when the duration seed is stale (GST-2193)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Refresh the free-suite duration seed from run 42691699570 (GST-2193)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Admit the eight test files the fresh seed records over 60 seconds as annotated exceptions (GST-2193)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep a numbered review finding's closing paragraph on that finding after a nested list (GST-2174)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep the nested-list join from lowering a separate finding or a CRITICAL item
The join that keeps a numbered native finding's closing paragraph now ends after that first paragraph, skips bold-led margin paragraphs and verdict lines, applies only to the native voice's items, and never lowers an item that names CRITICAL or P0. Each of these let a separate finding's Severity and Blocks: no lower a FIXABLE blocker to Low (GST-2174).
Review-Finding: review/specialist-batched#1 reader=testing
Review-Finding: review/red-team#2 reader=red-team
Review-Finding: review/red-team#3 reader=red-team
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Describe the numbered-finding nested-list rule and add its changelog entry
docs/review-policy.md now says how the collector keeps a native numbered finding's first closing paragraph after a nested list (GST-2174).
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
A rerun attempt's 100-job page took GitHub ~11 s for gbrain's 86-job Test runs, past the git proxy's ~10 s cutoff, so every rerun failed verification with api_error. Read 50 per page and, after a 5xx, reread the whole list once at 25. Any failed page still throws, so a partial job list never counts.
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Add failing tests: ready and land accept a stacked child's receipt reviewed against its open parent branch
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Refuse a receipt reviewed against another branch than the queue base at ready, land and train build (GST-2153)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Regenerate the doctor result anchors for the new receipt base mismatch code
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* fix: pre-landing review fixes
Train build now has a test that refuses a receipt reviewed against a parent branch.
Review-Finding: p1/specialist-batched#1 reader=specialist-batched
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* fix: pre-landing review fixes
Keep the base-contains-head diagnostic ahead of the receipt base check, normalize any remote's refs, and tell callers to merge the base in before the carry.
Review-Finding: v1/outside-voice#3 reader=outside-voice
Review-Finding: v1/outside-voice#1 reader=outside-voice
Review-Finding: v1/outside-voice#2 reader=outside-voice
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Say in the changelog entry that the base is merged into the child before the carry
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* List the receipt base mismatch refusal in the land refusal tables
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Leave the crew captain refusal table for a follow-up so this PR changes no agent instructions after review
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Make autoplan's rec and all replies apply every recommended plan change (GST-2139)
The Phase 4 reply table sent `all` to "approve as written" while gstack-gate
parse takes every recommendation, plan changes included, and it let `p1a`
approve with a user challenge still pending. The table now matches the gate:
`rec`/`all` is the one-word default and routes plan changes through B/B2,
a pending item approves nothing until the owner answers it, and p1 recommends
b when a recommended option changes the plan. A new test reads the table and
checks it against bin/gstack-gate parse for all, rec and p1a.
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* fix: pre-landing review fixes
Review-Finding: p1/red-team#1 reader=red-team
Review-Finding: p1/red-team#2 reader=red-team
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Rename the changelog fragment to the branch's fragment name
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Add failing tests for env.kv flagging a Python attribute read like self.client.session (#4209)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Read a Python or Dart attribute reference assigned with = as a name in env.kv (#4209)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep /autoplan working after a Claude Code 2.1.317 compaction (#4200)
On 2.1.317 a compact boundary's logicalParentUuid names a record the
session file never receives, and the preserved list ends with the last
record kept in place. The journal reader required that tail to equal the
logical parent, so the boundary became an orphan, every later record was
dropped, and each phase entry was refused on every retry.
The reader now accepts a preserved list whose last entry is the segment's
tail, and continues a boundary whose logical parent is absent from the
file from that tail. The 2.1.300 path is unchanged: when the tail is the
logical parent it still needs the boundary cycle, so an unflushed tail
supplies no ownership. A boundary that opens the file stays a resumed
fork's root, and lists naming no node of this session still continue
nothing.
Co-authored-by: vilano93 <244097767+vilano93@users.noreply.github.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Say the troubleshooting guide's compaction note covers Claude Code 2.1.317
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: vilano93 <244097767+vilano93@users.noreply.github.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Rebuild the decision snapshot from the merged log when two machines' syncs conflict (#4174)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep redacted decisions out of a rebuilt decision snapshot
The jsonl-append merge driver takes the union of base, ours and theirs, so a
log that lost lines on one side (--compact expunging a redacted decision)
would get them back. The snapshot rebuild now runs only when every base line
of the log is still on both sides; otherwise the merge aborts as before.
A conflicted snapshot when local edits are put back after the merge is now
rebuilt from the working-tree log instead of left at the merged version.
Review-Finding: p1/red-team#1 reader=red-team
Review-Finding: p1/red-team#2 reader=red-team
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Put the stash recovery path back to how main handles it
Rebuilding a conflicted decision snapshot after stash pop made a decision
that another machine redacted and compacted away consistent and publishable,
because stash pop merges the log as a union. Without the rebuild the
mismatched log and snapshot stay held, as on main. The redaction guard on
the merge-retry rebuild stays. A new test shows the brought-back decision
stays held and never reaches the remote.
Review-Finding: p2/outside-voice#1 reader=outside-voice
Review-Finding: p2/adversarial-native#1 reader=adversarial-native
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Fix seven test and eval checks that misread correct work or raced their neighbours (GST-2131)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Record the harness decision for the flaky-pass helpers (GST-2131)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Register the captured scrape fixture as free (GST-2131)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Point the whole codex-review excerpt at the checkout and stop counting refused or conditional defers (GST-2131)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Make the flaky-pass changelog entry read plainly and name lib paths (GST-2131)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Make the behind= trailer and rebase rules agree across gstack docs (GST-2136)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep the Capy skill inside its context budget after the behind= rewrite (GST-2136)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Skip the unset fix for a pin with a recorded reason, and add the changelog fragment (GST-2136)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Make autoplan's plan-approval gate item p1 so gstack-gate accepts it (#4181)
/autoplan Phase 4 told the agent to write a gate item with id `plan` and
to read replies `plana`/`planc`/`pland`/`plane`. gstack-gate only accepts
ids of a lowercase label plus a number, so render refused the gate and the
replies never parsed. Rename it to `p1` (the id lib/autoplan-gate.ts and
docs/unattended.md already use), lower the autoplan size pins by the 10
characters saved, and add a test that builds the gate from the template text
and runs it through bin/gstack-gate render and parse.
Co-authored-by: Eng. Mahmoud Bek ElNemer <83190847+ElNeMeR70@users.noreply.github.com>
Co-authored-by: Yuan S. <3579304+szsunyuan@users.noreply.github.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Re-ask autoplan gate replies that leave an approval pending (#4181)
Now that the /autoplan Phase 4 gate renders, its reply mapping is reachable.
gstack-gate parse accepts p1a while a User Challenge is still pending, and
treats all as rec, which takes plan-changing recommendations, yet the skill
sent both straight to A (approve as written). Re-present a reply that leaves
an approval item pending, and send a plan change chosen on a d<n> or uc<n>
item, all included, to B or B2. The paragraph is 3 characters shorter, so
the 12 autoplan size pins drop by 3.
Review-Finding: p1/outside-voice#1 reader=outside-voice
Review-Finding: p1/adversarial-native#1 reader=adversarial-native
Review-Finding: p1/adversarial-native#2 reader=adversarial-native
Review-Finding: p1/red-team#1 reader=red-team
Review-Finding: p1/red-team#2 reader=red-team
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Revert "Re-ask autoplan gate replies that leave an approval pending (#4181)"
This reverts commit b746d50dc5.
---------
Co-authored-by: Eng. Mahmoud Bek ElNemer <83190847+ElNeMeR70@users.noreply.github.com>
Co-authored-by: Yuan S. <3579304+szsunyuan@users.noreply.github.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Make the BROWSE_TUNNEL=1 auto-start share the browse tunnel start lock (GST-2117)
#4210 made ctx.tunnel.start single-flight, but the BROWSE_TUNNEL=1
auto-start in start() called startTunnel() directly, so a /tunnel/start
that overlapped it still opened a second ngrok session closeTunnel could
not reach. The lock now lives in startTunnel itself, which both callers
use. The unreleased changelog fragment names both paths.
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Split the browse tunnel changelog sentence at its colon (GST-2117)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Restore the landed tunnel changelog fragment and add this fix's own (GST-2117)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Teach crews to keep work owned and treat outside PRs as bug reports (GST-2077)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Name the credit range and who posts a verdict in the outside-PR docs (GST-2077)
Review-Finding: p1/red-team#2 reader=red-team
Review-Finding: p1/red-team#3 reader=red-team
Review-Finding: p1/red-team#4 reader=red-team
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Name the changelog fragment after the branch and mention the credit range (GST-2077)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Check fork authors before adding contributor credit (GST-2077)
Review-Finding: p1/red-team#1 reader=red-team
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Describe contributor PRs as rewritten fixes in the harness options table (GST-2077)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
Two /tunnel/start calls that both saw the tunnel off each opened an ngrok
session; the daemon kept only the second, so the first stayed public and
closeTunnel could not reach it. ctx.tunnel.start is now single-flight, so a
concurrent caller shares the in-flight start and its URL.
A start that failed after ngrok.forward() reset only tunnelListener, leaving
tunnelActive, tunnelUrl and tunnelServer set, so the daemon never idled out.
The catch now clears all four.
Tests found and first written by GEVA-449's pstack verification.
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep quoted file names with secrets in review (GST-1451)
The review collector now decodes git's C-quoted patch headers, so the secret-line keep rule reads the added lines of a file whose name holds a quote, backslash or control byte. reviewExclude also keeps a matched file the collector holds no lines for, as unread.
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Type the expected kept reasons in the quoted-name test
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Skip the quoted-name tests on Windows, which cannot create those names
Review-Finding: p1/outside-voice#1 reader=outside-voice
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep deferred review findings on the PR instead of filing issues
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Repin the ship review section after the deferred-findings lead edit
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep the deferred record's severity as the plain label
The reviewer's stated severity and Blocks text now sit in their own field, so a review-log row built from a deferred record keeps a severity the ledger can rank. The PR body renders the same line as before.
Review-Finding: p1/adversarial-native#2 reader=adversarial-native
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Say deferred findings stay on the PR in the unattended docs
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Let the docs audit repair a badly shaped final line without using an attempt
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep the audit id and outcome out of docs audit format repairs
A missing audit id or a missing or unknown status is now a content failure, a --repair-of that names a non-repairable output no longer uses up the audit's one repair, and the re-ask goes to a fresh agent with no tools instead of the writer child.
Review-Finding: p1/specialist-batched#1 reader=testing
Review-Finding: p1/specialist-batched#2 reader=testing
Review-Finding: p1/red-team#1 reader=red-team
Review-Finding: p1/red-team#2 reader=red-team
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Treat repeated identity keys and malformed outcome lists as failed docs audits
A repeated audit_id or status key, and a files_updated or blockers list that is present but malformed, are now content failures, so a format repair can no longer erase a reported blocker or choose the outcome. A missing list stays repairable.
Review-Finding: v1/outside-voice#1 reader=outside-voice
Review-Finding: v1/outside-voice#2 reader=outside-voice
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
@types/bun 1.4.3 -> 1.4.4, @huggingface/transformers 4.3.1 -> 4.3.2 and
marked 18.1.1 -> 18.1.3 (package.json and bun.lock), docker/login-action
4.6.0 -> 4.7.0 and github/codeql-action/upload-sarif 4.39.0 -> 4.39.1,
each pinned to the full SHA of its upstream tag. Rebuilt on current main
from Dependabot's #3488, #3489 and #3890.
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Stop flagging test run numbers in PR descriptions as phone numbers (GST-2047)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Let approval import record autoplan run ids that look like phone or card numbers (GST-2047)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Read versions beside a VERSION file or a semver as versions, not public IPs (GST-2047)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Say pre-merge checks in the changelog entry (GST-2047)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep the phone-number check on run-id fields that hold other text and on run id names (GST-2047)
Review-Finding: p1/red-team#1 reader=red-team
Review-Finding: p1/specialist-batched#1 reader=security
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Exempt only 11 to 20 digit run numbers in a ship summary block from the phone-number check (GST-2047)
Review-Finding: v1/adversarial-native#2 reader=adversarial-native
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep local paths out of the approvals table and add render --redacted
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Label local paths in the Asked and Still pending columns too
pathLabels now covers file:// URIs, ~user, $HOME, %USERPROFILE%, UNC and Drive-relative paths, and the approvals table runs it on Asked and Still pending as well as Covers. Exact words stays verbatim.
Review-Finding: p1/red-team#1 reader=red-team
Review-Finding: p1/specialist-batched#1 reader=security
Review-Finding: p1/specialist-batched#2 reader=testing
Review-Finding: p1/red-team#2 reader=red-team
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Label paths glued onto a colon or dash, and file: in any case
Only a :// scheme prefix now keeps a slash path unlabelled, so ledger:/home/..., user@host:/home/... and run-/home/... get file-name labels, and file:/ and FILE:/// paths match. Granted fix-check a-5b68cd273658.
Review-Finding: p2/adversarial-native#1 reader=adversarial-native
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Stop the focused-test tag from keeping run data in review (GST-2006)
RH-1's skip, focus and removed-declaration checks now read only files a test runner can execute: source files by extension, extensionless scripts, test-folder or project-glob files that are not data, and files renamed from one of those. English like "don't fit (" in a results .jsonl no longer tags the file and keeps it in the reviewers' diff under review_exclude_paths (harness-evals#333).
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep RH-1 on every file except data and logs
Review found that the code-extension allowlist dropped RH-1 from an HTML browser suite, unlisted languages, and a test-floor file renamed to data. RH-1 now skips only .json, .jsonl, .ndjson, .csv, .tsv, .log and .out files that were not renamed from another kind of file.
Review-Finding: p1v/outside-voice#1 reader=outside-voice
Review-Finding: p1v/outside-voice#2 reader=outside-voice
Review-Finding: p1v/targeted#1 reader=targeted
Review-Finding: p1v/adversarial-native#2.1 reader=adversarial-native
Review-Finding: p1v/adversarial-native#2.2 reader=adversarial-native
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep the scroll selector when --wait is absent
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Skip a PATH browse that links to another program such as xdg-open
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Give each project its own browse slug on installs without ~/.claude
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep a question marker when its signal hint is malformed
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Read the lowercased question id in the never-ask preference hook
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep a marker's tail in the one-way door check
Review-Finding: p1/targeted#1 reader=targeted
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep a marker's tail apart from the next word in the one-way door check
Review-Finding: p2/outside-voice#1 reader=outside-voice
Review-Finding: p2/adversarial-native#1 reader=adversarial-native
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Say where old domain skills live and narrow the marker note in the changelog
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Say how the browse daemon picks a project's domain skill file
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Add failing tests for fix-check chains whose earlier links found issues (GST-1973)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Count fix-check chains whose earlier links found issues a later clean link read (GST-1973)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* List the new follow-up review test with the other release-script tests in the test-only ratchet (GST-1973)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Match finding paths written with a leading ./ and test the remaining refusal guards (GST-1973)
Review-Finding: p1b/specialist-batched#1 reader=testing
Review-Finding: p1b/red-team#1 reader=red-team
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Name the changelog entry after the branch and say it in plain words (GST-1973)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Reword the changelog entry in plain words without tracker codes (GST-1973)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Refuse a follow-up review chain when no later clean review read an earlier finding's file (GST-1973)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Say in the changelog entry that a finding no later clean review read now refuses the chain (GST-1973)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Report free-suite results to Trunk Flaky Tests for detection only (GST-1859)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Say that test files that time out or crash do not reach Trunk
Review-Finding: p1/red-team#2 reader=red-team
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep a failed JUnit upload from failing a free-suite shard, and correct what the docs say about timed-out children
Review-Finding: p11/outside-voice#1 reader=outside-voice
Review-Finding: p1fc/fix-check#1 reader=fix-check
Review-Finding: p1fc/fix-check#2 reader=fix-check
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Say the Trunk uploader job also sees a secret in Free Tests
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Add failing tests for repo-declared docs-only paths (GST-1974)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Let a repository name its own docs-only paths in its base ship policy (GST-1974)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Catch relative directory readers, missing docs tests and readers the base gained in the docs-only class
Review-Finding: review/specialist-batched#1 reader=security
Review-Finding: review/red-team#1 reader=red-team
Review-Finding: review/red-team#2 reader=red-team
Review-Finding: voices/adversarial-native#1 reader=adversarial-native
Review-Finding: voices/adversarial-native#3 reader=adversarial-native
Review-Finding: voices/adversarial-native#5 reader=adversarial-native
Review-Finding: voices/outside-voice#1 reader=outside-voice
Review-Finding: voices/outside-voice#2 reader=outside-voice
Review-Finding: voices/outside-voice#3 reader=outside-voice
Review-Finding: voices/outside-voice#4 reader=outside-voice
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Say in the changelog that a missing docs test or a new base reader keeps the full checks
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Describe the new directory-reader forms, the missing docs test refusal and land's base-reader check
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Make the autoplan docs say what lean and full runs actually load (GST-1985)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep the autoplan description at its size pin and say when the eng checklist is sent
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Let review exclusions handle big results PRs (GST-1967)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep the value rule in review-exclude so the pinned review-risk rules stay at their replayed bytes
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Reword the changelog entry so its credential sentence reads plainly
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Let pull requests that only change human-facing docs skip the test suites (GST-1951)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Quote the docs test list in the docs lane so shellcheck passes
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep the Windows smoke check green on docs-only pull requests and update the workflow contract tests
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep the merged test check when the base gained a reader of a changed doc, hold the land policy key until a release reads it, and register two more docs reader tests
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Check links in untracked docs and docs with quoted or spaced names, and report an undecodable anchor instead of crashing
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Point the flaky-exception anchor back at its section and test the link checker's exit codes
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Stop pregate flagging real files whose names start with tmp or scratch (GST-1944)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Make the stray-name test safe on case-insensitive disks and record the tmp-word tradeoff
Review-Finding: p1/specialist-batched#1 reader=testing
Review-Finding: p1/red-team#2 reader=red-team
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Describe the new scratch-name rule in the PREGATE_STRAYS troubleshooting entry
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep crew threads quiet and let busy Captains hand CI watching to one merge thread (GST-1941)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep crew threads quiet and let busy Captains hand CI watching to one merge thread (GST-1941)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep the READY note with its list and cap CI reruns at one in the Captain guide
Review-Finding: capy/skill/gstack/captain.md:162:completeness reader=core
Review-Finding: capy/skill/gstack/captain.md:157:completeness reader=core
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Stop macOS EPERM marking a finished QA capture as incomplete
On macOS the cleanup kill of an exited command's process group can get
EPERM while the group is being torn down, and the capture then recorded a
successful command as incomplete with exit code 2. On EPERM, check the group
directly (kill with signal 0, then a ps listing that ignores zombie and
exiting entries) and treat it as gone only when no live member remains.
A live member or a failed check keeps the fail-closed result.
Co-authored-by: vilano93 <244097767+vilano93@users.noreply.github.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep the EPERM group check to macOS so Linux still fails closed
Review-Finding: p1/red-team#1 reader=red-team
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Count exiting macOS group members as gone after a refused group kill
On macOS, ps shows a process that is exiting with a trailing E flag
(for example RE or UE), so a check on the leading letter alone read a
mid-exit member as live and kept the capture incomplete. Count a member
as gone when its state starts with Z or carries E; X means traced on
macOS, so it stays live. Linux is unchanged and still fails closed.
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Document the macOS EPERM exception in QA deadline cleanup
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: vilano93 <244097767+vilano93@users.noreply.github.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Count an autoplan phase report that starts after a preface sentence (GST-1916)
Claude Code can journal the prose between tool calls as a paraphrased
narration summary. A Phase 1 report that opened with a check sentence came
back as "I checked the packet... Phase 1 is complete with 9 native issues and
no outside review, so the work moves on to Phase 2.", and the publication
guard denied the next phase with publication_missing because it counted a
declaration only at the start of a line (#4155).
- A declaration that starts a sentence inside a plain prose line now counts.
Sentences inside double, curly or single quotes or inline code, and list,
heading, table and quote lines, still don't.
- A "with" result that mentions "no" after its counts is affirmative; a
result that opens with "no" is still rejected.
- New test replays the captured summary as a narration-journaled report
through the guard, with quoted, negated and other-phase controls.
Co-authored-by: Marius Storm-Olsen <1500+mstormo@users.noreply.github.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Reject an autoplan phase report that follows a same-line example introduction (GST-1916)
Review-Finding: review-p1/specialist-batched#1 reader=red-team
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Mention same-line example introductions in the autoplan changelog entry (GST-1916)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: Marius Storm-Olsen <1500+mstormo@users.noreply.github.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Fix four flaky test causes found in the 2026-11-08 harvest (GST-1912)
- gbrain prune test: stub pgrep so a parallel test's 'gbrain autopilot' process cannot refuse the removal (Free Tests 42448270551 shard 3).
- third-party-actions: a lettered option that marks the step blocked or waiting on the user is the defer option (gate 42431616220, tpa-present).
- gstack-hook-check: a Bun killed by a signal makes the check unavailable (exit 3) instead of failing every hook (Free Tests 42327570347 shard 8).
- floor actor: bind a cropped Write overwrite to the sole pending owned Write, checking every visible diff row (census 42252633016, plan-ceo-finding-floor).
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Move the overwrite permission cases into the file-permission owner test
The test-of-test ratchet fails a new file that imports only test code. The cases now live in test/plan-count-file-permission.test.ts, the ratchet's suggested owner, and the fixture map points there.
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep checking shell hooks when Bun crashes during the hook check
A Bun killed by a signal used to end the check before the shell parse and merge-marker scan. Those checks now still run: a shell failure exits 1 and setup refuses that hook, and otherwise the check exits 3 as before with TypeScript hooks left unverified.
Review-Finding: p1/red-team#1 reader=red-team
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep the hook failures Bun reported before it crashed
When Bun printed a fail row for one entry and was then killed by a signal, the check dropped that row and exited 3, so setup registered a hook it knew was broken. Rows Bun printed now count: a fail row fails its hook, an ok row adds its imports to the marker scan, and only entries with no row are unchecked. Exit 3 applies only when an entry has no row and nothing else failed.
Review-Finding: v2/outside-voice#1 reader=outside-voice
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Leave the third-party-actions defer check as it is on main
The widened defer rule counted a refusal such as "I will not mark this step as blocked on you" as a defer option, which loosens the paid consent gate. This change returns test/helpers/third-party-actions.ts and its test to origin/main; that flaky case moves to a separate fix.
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Describe only the three repairs this branch still ships in its changelog entry
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Load the Capy crew skills only on Capy machines (GST-1833)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Add changelog entry for Capy-only crew skills (GST-1833)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Move the Capy-only crew tooling into crew/ and run its tests only when needed (GST-1833)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Test that the Capy launch file puts crew/bin on PATH
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Count any ~/.capy as a Capy machine, so a Mac running Capy keeps /captain
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Fix the tests and fixtures that broke when crew tooling moved into crew/
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Plan every crew test when a pull request's merge-base can't be read
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Add a test that checks each crew-only module left the root tree for crew/
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Run crew tests on pull requests that change bin/ commands or workflows they drive
Review-Finding: p1/specialist-batched#1 reader=testing
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep running every crew test on every pull request; the skip comes in a later change
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep copied installs and older release bisects working after the crew/ move
Setup now copies crew/lib into every host's runtime root beside bin and lib, so Windows installs (which copy instead of linking) can still start gstack-ship-receipt, gstack-pregate and the other commands that import it. The release bisect runs the train-gate script from wherever the checked-out commit has it, so a bisect over commits from before the move still gets a verdict.
Review-Finding: p1/outside-voice#1 reader=outside-voice
Review-Finding: p1/outside-voice#2 reader=outside-voice
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Say in the changelog that Windows installs keep working after the crew/ move
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Point the harness guide's merge-queue rows at crew/lib/ where those modules moved
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Point the plain-title test at the wave module's new home in crew/ (GST-1833)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Give a clear error when Chromium's sandbox can't start, and make design-build preflight launch the browser (GST-1906)
On Linux hosts that deny unprivileged user namespaces (stock Ubuntu 23.10+ under AppArmor, non-root Docker), Chromium aborts with "No usable sandbox!". browse now fails with one error naming both fixes (allow user namespaces, or GSTACK_CHROMIUM_NO_SANDBOX=1 for trusted pages only) with Chromium's line below it, and never turns the sandbox off on its own. Docker's /.dockerenv now counts as a container, like CONTAINER. gstack-design-build preflight opens and closes one tab in the browser it found and stops with code=BROWSER_SANDBOX and the same error when that launch fails.
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Reword the Chromium sandbox changelog entry (GST-1906)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Stop /ship asking for a money review of files the review policy leaves out (GST-1900)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* chore: changelog and TODOS
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Stop the secrets check flagging code that reads an environment variable (GST-1894)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep the new secrets test rows out of the line-start credential shape
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Give the fixture-label exemption only to files named like tests
A production file such as ab-test.ts or load-test.ts matched the test-file pattern, so a weak real password in it stopped being flagged. The pattern now matches only the documented forms: *.test.*, *.spec.*, *_test.* and test_*.py.
Review-Finding: p1/red-team#1 reader=red-team
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep flagging a hardcoded fallback in backticks or on the next line
The env-read exemption only looked for a quoted fallback on the same line, so token = opts?.env?.API_TOKEN ?? followed by a backtick literal, or by a literal wrapped onto the next line, passed the secrets check. Backtick literals now count, and the next line is read when the expression continues there. The changelog now says plainly that an ALL-CAPS literal used as the variable name still passes.
Review-Finding: voices/outside-voice#1 reader=outside-voice
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Exempt an env read only when its line has no fallback, and stop reading the next line
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Say in the changelog that quoted lowercase env names now fail the secrets check
Review-Finding: p1/red-team#1 reader=red-team
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep reading an older receipt at the top of a PR body, and leave quoted receipt examples in place
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Put a receipt posted on its own inside the PR body's Agent notes, so the new one is the one read
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Read the PR body from the event file so a very long body cannot break the title check
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Make the Plain title check required before gstack merges a PR
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep the PR's opening paragraph when a posted receipt moves into the Agent notes
Review-Finding: v1/adversarial-native#1 reader=adversarial-native
Review-Finding: v1/outside-voice#1 reader=outside-voice
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Say exactly when an older receipt at the top of a PR body is still read
The older receipt is read only while it still opens the body. A receipt quoted in the readable part is never read only when the body has an Agent notes section. The example policy now lists the Plain title check.
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Add the changelog entry for the required Plain title check
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Hide tokens in URLs that do not parse or that hold punctuation
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Add a build receipt and a fixture preview server so a design state whose API call went unanswered is never reported finished
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Register the preview fixtures as free so editing them does not rerun every paid test
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Add /design-review --build, an opt-in loop that lints, screenshots and fixes UI while you build it
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Recommend the design build loop for UI work in the lane card, routing rules, skills guide and design-html
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Let the design review skill grow by the opt-in build mode's routing line and pointer
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Make the design build loop work when an agent follows it step by step
A real run of /design-review --build turned up four problems:
- Each bash block runs in a fresh shell, so later blocks lost $RUN and the
helper path, and the first block never set the project slug or state root.
Blocks now stand alone and later ones use the printed run path.
- The lint command it printed was a bare name that is not on PATH; it now
prints the full path to the bundled lint.
- gstack's own browser can only save screenshots under /tmp, so the section
says to save where the browser can write and copy the file into place.
- The --build pointer sat before the design hard rules, which every section
must follow; it now comes right after them.
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Hold the design-review size ratio to the granted 1.357
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Let the free checks find the preview server's test when the server file changes
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep the design build loop honest about stale API misses and pages that are not the preview
A missing fixture route that was fixed before the final pass no longer blocks the receipt, an API call that arrived with no scenario now does, preflight only promises fixtures-only mutations when every page carries the preview script, and a bad path escape or a fixture with a bad status no longer stops the preview.
Review-Finding: p1/specialist-batched#1 reader=state-machines
Review-Finding: p1/specialist-batched#2 reader=security
Review-Finding: p1/red-team#1 reader=red-team
Review-Finding: p1/red-team#2 reader=red-team
Review-Finding: p1/red-team#4 reader=red-team
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Make the design build loop fail closed on any missed API call and only trust a page the preview actually serves
Any API call the preview could not answer during a run now keeps the receipt incomplete, and its fix line says to add the route, restart the preview and start again from preflight. Preflight checks every local page whatever hosts were granted, asks the preview's own probe route instead of looking for a marker in the HTML, and treats a redirect as not the preview. The Vite plugin now applies to the dev server only, so built pages never carry the preview script.
Review-Finding: p2v/outside-voice#3 reader=outside-voice
Review-Finding: p2/red-team#1 reader=red-team
Review-Finding: p2v/outside-voice#1 reader=outside-voice
Review-Finding: p2v/outside-voice#2 reader=outside-voice
Review-Loop-Exception: a-0aec1d2672bd
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Link the frontend preview guide from the README and note the preview check in the changelog
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Fix a test type error the CI type check caught in the preview probe test
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Fix the five free-suite failures in the design build loop
Run the slug helper through the shared script launcher, move the preview
server tests into the design build test that imports the product code, let
the single-source check skip generated host copies, write the build-loop
commands without free-text placeholders (trimmed so the skill stays under its
size cap), and add the build routing line to the onboarding fixture.
Review-Loop-Exception: a-b2e2abe1c646
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Document how to waive findings in the design build loop and say which requests the preview log records
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep the design build receipt from calling waived, edited or proxied runs clean
The receipt now reads waivers from the lint's grade, hashes every file in a touched folder, refuses an inventory that changed after verify-start, and finds the preview's record of unanswered calls in its default state folder. The preview answers a path the app's own dev proxy would forward with a recorded 501, and refuses a state folder that is a symlink or another user's.
Review-Finding: p1/red-team#1 reader=red-team
Review-Finding: p1/red-team#2 reader=red-team
Review-Finding: p1/red-team#3 reader=red-team
Review-Finding: p1/red-team#4 reader=red-team
Review-Finding: p1/specialist-batched#1 reader=testing
Review-Finding: p1/specialist-batched#2 reader=security
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Give the new receipt tests the state mode the inventory type requires, and say the preview does not stop proxied WebSocket upgrades
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Stop a catch-all dev proxy, an unrecorded preview and passwords with @ from slipping past the design build checks
An empty server.proxy key now gets the recorded 501 too. The preview's probe names its state directory, so the receipt reads a --state or --port 0 preview's record without --preview, and a preview that keeps no record makes the run incomplete. Text redaction drops userinfo up to the last @ and redacts percent-encoded secret parameter names.
Review-Finding: s11/outside-voice#1 reader=outside-voice
Review-Finding: s11/targeted#1 reader=targeted
Review-Finding: s11/adversarial-native#3 reader=adversarial-native
Review-Finding: s11/targeted#2 reader=targeted
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Build the userinfo redaction test's URL from parts so the credential scan does not read it as a real login
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep reading a stopped preview's default record when another origin's preview still answers
The receipt now reads the union of --preview, the state directory each answering preview names, and the default gstack-preview-<port> directory for every loopback state origin that did not answer the probe. servedStates reports which origin each answer came from.
Review-Finding: fc3/fix-check#1 reader=fix-check
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* List the design build receipt, its library and the preview template in the project structure guide
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Stop /ship stalling on branches that merged in thousands of data files (GST-1855)
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Still spot scripts in data files when git grep is set to fixed strings
Review-Finding: p1/red-team#1 reader=red-team
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Document that a review plan run by itself also stops at its time limit
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Add a check that keeps tracker codes and internal jargon out of PR titles and branch names
gstack-plain-title checks the readable front of a title, branch name, PR opening or changelog entry against the repository's .gstack/plain-titles.json; one trailing issue code stays allowed. A Plain title workflow runs it on every new pull request, and /ship's push refuses a never-pushed branch whose name fails it.
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Move the PR body's machine-read lines into a closing Agent notes section
The ship receipt, its runs block and the Plan line now live in one collapsed section at the end of the body, and every reader looks there first while still accepting the old receipt-at-top layout. gstack-post refuses to post a title the plain-title check refuses, and land uses the checked PR title as main's commit title.
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Teach /ship to write plain titles, PR openings, commit messages and changelog entries
The README now lists what gstack is for, so a PR opening can name the goal it serves.
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep tests, size budgets and generated files in step with plain titles and Agent notes
Copy lists that install result-code modules now include the plain-title codes, the ship goldens and size ledger are refreshed, the new workflow is classified, and the pinned prose tests read the new title and body layout.
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Record the test-profile list change as non-behavioral for the shared test harness
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep text people add below the PR notes, and fix three smaller review findings
The notes section now ends at its own closing tag, so a collapsible added after it survives the next post. The opening check splits the body the same way the notes reader does. Wave PR titles no longer name the area, which the plain-title check could refuse. The Capy line in the PR-body steps now puts Approvals in the notes.
Review-Finding: p1/specialist-batched#1 reader=testing
Review-Finding: p1/red-team#1 reader=red-team
Review-Finding: p1/red-team#2 reader=red-team
Review-Finding: p1/red-team#3 reader=red-team
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Mask credentials in refused-title lines and keep older PRs' plan lines readable
A refused title's printed match is masked by the secret scanner first. The Plan line reader falls back to the whole body when the Agent notes section has none, so a body written before this change keeps its plan when its first receipt lands. The notes section also ends at its own closing tag when its opener is split or marked open, or when a collapsible inside it is left unclosed.
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Print a fixed placeholder when a credential in a refused title cannot be masked
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Keep a plain-title policy edit from scheduling every paid eval
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Record the paid-eval list change as non-behavioral for the shared test harness
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
* Recheck the PR title on every edit so a base change cannot turn a refused title green
Review-Finding: s11/targeted#1 reader=targeted
Review-Finding: s11/adversarial-native#1 reader=adversarial-native
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
---------
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
