Commit Graph
546 Commits
Author SHA1 Message Date
montjeffrey e56624d8f1 test(clv2): catch 3.8 and 3.12 f-string forms in the compat guard
ast.parse(feature_version=...) on current Pythons goes no lower than the
3.7 grammar, and it accepts f-string forms that older interpreters reject:
the self-documenting {x=} (3.8) and, from 3.12, reused quotes, backslashes
and comments inside a replacement field. ast does not record these, so
the guard now reads each f-string's own source text. The sample file
proves both kinds are flagged and that 3.6-valid fields ({n == 1},
{n!r:>4}, escaped {{n=}}) are not.
2026-11-04 11:45:32 -05:00
montjeffrey e41a4a0492 feat(clv2): tell instinct-cli users on an end-of-life Python to upgrade
instinct-cli keeps running on old interpreters, but people on one should
know they no longer get security fixes. On a Python older than 3.11 (3.10
reached end of life in October 2026) the CLI now prints one line to
stderr after parsing its arguments:

  [instinct-cli] Python 3.8 is past end of life. instinct-cli still runs
  on it; upgrade to Python 3.11 or newer for security fixes.

The command itself runs as before. The notice stays out of automated
output: `prune --quiet`, which observer-loop.sh runs, skips it, and
CLV2_SKIP_PYTHON_EOL_NOTICE=1 turns it off for anyone who keeps an old
Python on purpose. The threshold is one constant, _SUPPORTED_PYTHON, to
raise as versions age out. Importing the module prints nothing.
2026-11-04 11:30:07 -05:00
montjeffrey 6053e1a11b fix(clv2): keep instinct-cli running on Python 3.6 and newer
The hooks fall back to any Python 3 that runs, so users on an older
interpreter (RHEL 8's 3.6, for one) still reach instinct-cli. It failed
there at import: builtin generics such as list[dict] in signatures are
evaluated when the function is defined, which raises TypeError before
3.9, and str.removeprefix is 3.9+.

- Quote the builtin-generic annotations. Type checkers read them the
  same way, and they stay valid under 3.14's deferred annotations.
- Strip the byte-order mark with startswith and a slice.
- Open the registry lock file with an explicit UTF-8 encoding.
- Say in the module docstring that 3.6 is the floor and what keeps it.

The new test parses the script with the oldest grammar ast offers and
rejects what a grammar check cannot see: unquoted builtin generics in evaluated
annotations, `from __future__ import annotations`, and 3.7+ methods and
keywords (capture_output, text, removeprefix and similar). A sample file
proves it catches each of those. It also runs import, status, export,
evolve, projects and prune under -X dev -X warn_default_encoding
-W error, so any deprecation or default-encoding warning on the current
Python fails the build. CI runs only current Pythons; this is a static
guard, not a run on a real 3.6.
2026-11-04 11:28:28 -05:00
montjeffrey 3f56735bc3 fix(clv2): decode instinct-cli git output as UTF-8
git prints repository paths and remote URLs as UTF-8 on every platform,
but instinct-cli read them in text mode without an encoding, so Python
decoded them with the locale encoding. On Windows that is the ANSI code
page (usually cp1252): a non-ASCII repo path came back garbled ("café"
became "café"), a byte cp1252 cannot map crashed detect_project with
UnicodeDecodeError, and a non-ASCII remote URL hashed to a different
project id than the URL itself does.

The three git calls now decode as UTF-8 with errors="replace". They are
spelled with stdout/stderr=PIPE instead of capture_output/text, which
are Python 3.7+.

The new test forces a cp1252 locale encoding, so it reproduces off
Windows too. All three cases fail without the fix and pass with it.
2026-11-04 11:16:13 -05:00
haelyra 60053622fe maint: integrate reviewed contribution #3791 2026-10-29 11:34:19 -07:00
haelyra a98821fff9 maint: integrate reviewed contribution #3779 2026-10-29 11:34:15 -07:00
haelyra 2ecb5fcc04 maint: integrate reviewed contribution #3752 2026-10-29 11:30:40 -07:00
haelyra 6fa858fbdf maint: integrate reviewed contribution #3803 2026-10-29 11:27:56 -07:00
haelyra 023acf21d0 maint: integrate reviewed contribution #3794 2026-10-29 11:27:56 -07:00
haelyra 6a3c45ac7f maint: integrate reviewed contribution #3601 2026-10-29 11:25:27 -07:00
haelyra ff147d751f maint: integrate reviewed contribution #3593 2026-10-29 11:25:25 -07:00
haelyra f77db133ac maint: integrate reviewed contribution #3614 2026-10-29 11:25:23 -07:00
haelyra a60c58ab42 maint: integrate reviewed contribution #3780 2026-10-29 11:25:19 -07:00
haelyra d2c71fd0e9 maint: integrate reviewed contribution #3588 2026-10-29 11:25:18 -07:00
haelyra 09e0eb7d7f maint: integrate reviewed contribution #3591 2026-10-29 11:23:24 -07:00
minteeqahmed-png ef82fcb130 fix(scripts): strip unquoted YAML comments and preserve mixed body line endings 2026-10-29 21:49:53 +05:00
minteeqahmed-png fb4683317f fix(scripts): harden gemini-adapt-agents for CRLF, unknown flags, and block tool lists 2026-10-29 20:56:00 +05:00
longankang 8e4a6c03bf fix: validate Codex hook references in plugin cache
Collect string-valued manifest.hooks as a file reference so missing files, directories, and paths outside the cache fail validation. Add regression coverage for valid and invalid hook references while preserving non-string and absent-hook behavior.

Closes #3786
2026-10-29 15:30:33 +08:00
Harry Chen 24bed7165e test(work-items): admit older ready tasks past newest 100 entries 2026-10-29 12:37:02 +08:00
Harry Chen 69a46925ee test(work-items): cover automatic ready-only pickup 2026-10-29 12:22:28 +08:00
Harry Chen 2948adf3af test(work-items): cover legacy in-progress claim compatibility 2026-10-29 12:14:57 +08:00
Harry Chen 49f3329222 test(work-items): cover explicit ready claim admission 2026-10-29 12:01:02 +08:00
Shubham Padkonde 1d77da5be7 fix(work-items): leave imported issues available for native claims 2026-10-29 09:22:33 +05:30
Ali Arslan 4ebe4c8bc3 fix(tests): skip the inert OpenCode stub when loading plugin factories 2026-10-28 22:50:14 +05:00
Ali Arslan 6b11ffeb7d fix(tests): assert the inert OpenCode stub as a V2 definition object 2026-10-28 22:48:50 +05:00
Yann Roberto f1f542044c fix(install): accept all supported component families in config 2026-10-28 16:58:35 +08:00
Shubham Padkonde a75cdc322b test(work-items): verify claim preservation and report suite totals 2026-10-28 13:10:12 +05:30
Shubham Padkonde b805cdfae8 fix(work-items): preserve native ownership during GitHub sync 2026-10-28 12:52:50 +05:30
Muhammad Zubair 0618bdc7a8 test(claw): keep the askClaude tests from spawning a real claude CLI 2026-10-27 23:58:20 +05:00
Dante 49f8893dd3 fix(codex): bind cache rerun guidance to active plugin 2026-10-27 12:27:44 +08:00
Dante dd6f58d3d4 fix(codex): recommend native cache recovery in checker 2026-10-27 08:34:42 +08:00
Alan Vo 5701a4f2e1 fix: respect CLAUDE_CONFIG_DIR with WSL Windows-path conversion
(cherry picked from commit aaacbc7592)
2026-10-20 17:40:58 -04:00
Eneko Ruiz 65f7dfb735 fix: reject missing session database and limit values
(cherry picked from commit 299899941d)
2026-10-20 17:40:57 -04:00
Gaurav Dubey e4666abb01 fix(opencode): ship @opencode-ai/plugin as a runtime dependency
The compiled OpenCode tools import `tool` from `@opencode-ai/plugin/tool`
at module top level, and the hooks plugin imports those tools statically.
The package was only a devDependency, so production installs (OpenCode's
npm plugin cache, `npm install --omit=dev`) could not resolve it and the
whole plugin failed to load.

Move it to `dependencies` and add a build-opencode regression test that
fails when the built bundle imports a package not declared as a runtime
dependency.

Fixes #3543

(cherry picked from commit 106024070e)
2026-10-20 17:38:25 -04:00
Eneko Ruiz 5367671a22 fix: reject missing session inspection option values
(cherry picked from commit a6619f5efd)
2026-10-20 17:38:13 -04:00
KET ⚡ 72cf932195 fix(worker): clarify invalid task file diagnostics 2026-10-16 22:29:18 +05:30
Gaurav Dubey 159e389927 fix(dashboard): log when the browser cannot be opened automatically
openBrowser() captures the launcher's async error internally, so a missing
xdg-open/open left headless users with no hint why no browser appeared.
Wrap the injected spawn in the entrypoint to report the failure on stderr
with the dashboard URL; the shared platform-launch helper is unchanged.
2026-10-16 08:54:36 +00:00
Gaurav Dubey 405610bc29 fix(dashboard): keep the web dashboard running when no browser launcher is installed
The auto-open in the dashboard-web.js entrypoint spawned xdg-open/open
without an 'error' listener. When the launcher is missing (headless
Linux, containers, many WSL setups) spawn reports ENOENT asynchronously,
the try/catch cannot see it, and the unhandled 'error' event kills the
process right after the banner is printed.

Use the shared openBrowser() helper from scripts/lib/platform-launch.js,
which already handles the child's error event and unrefs the launcher,
as plan-canvas.js and control-pane.js do.

Fixes #3600
2026-10-16 04:37:11 +00:00
KET ⚡ dc42aa5409 fix(worker): reject non-file task inputs 2026-10-15 22:59:39 +05:30
azank 868ac1e894 test(hooks): build the session-start test env without mutating it
AGENTS.md asks for new objects rather than mutation. Leave NODE_PATH out
with a rest destructure, as pre-bash-tmux-reminder.test.js does, instead
of deleting it from the env after building it.
2026-10-15 19:21:38 +05:00
azank 218f4ae85f test(hooks): isolate the no-deps session-start test from NODE_PATH and the context cap
A NODE_PATH that exposes js-yaml let the regression pass the test, and an
inherited ECC_SESSION_START_MAX_CHARS=0 made it fail falsely. Drop
NODE_PATH from the child environment and pin the cap to 8000, as
session-start-alias-unavailable.test.js does.
2026-10-15 19:08:53 +05:00
azank 1d704e75e3 fix(hooks): load js-yaml lazily so session-start runs without node_modules
The merge of #3470 moved the Flutter check in scripts/lib/project-detect.js
onto js-yaml with a top-level require. session-start.js loads that module
on every SessionStart, and plugin installs have no node_modules (#2822),
so the hook now exits 1 with "Cannot find module 'js-yaml'" in every
project, not only Dart ones.

Require js-yaml inside hasFlutterSdkDependency instead. When it is
missing, the existing catch returns false: Dart is still detected from
pubspec.yaml and only the Flutter refinement is skipped. Installs that
have their dependencies behave as before.

Add a case to the no-node_modules plugin install test that runs
session-start.js against a pubspec.yaml project. It fails on main.
2026-10-15 16:38:43 +05:00
Yann Roberto 932aa393c1 fix: prevent GitHub work item identity collisions 2026-10-15 18:50:54 +08:00
haelyra 70f571e205 test: drain synthetic scanner input before reporting status 2026-10-11 20:23:42 -04:00
haelyra 87c136fcae fix: integrate strict issue targeting and Windows Python discovery 2026-10-11 15:07:29 -04:00
haelyra c7a467061b Merge remote-tracking branch 'origin/pr-3523' into steward/late-arrivals-20261011 2026-10-11 15:00:38 -04:00
akushonkamen 3cf0e709d5 fix(github-coordination): reject malformed issue numbers instead of truncating them
normalizeIssueNumber parsed with Number.parseInt, which stops at the
first non-digit ("42oops" -> 42, "42.9" -> 42, "4e2" -> 4) and loses
precision past 2^53 (9007199254740993 became 9007199254740992). The CLI
shares this helper for --issue, --limit, and the positional issue number,
so malformed values silently targeted the wrong issue before any integer
guard downstream could notice.

Validate the entire trimmed decimal token and require
Number.isSafeInteger so malformed or out-of-range values throw
"Invalid issue number" during argument parsing, before the state store
is opened or GitHub is contacted. Whitespace, leading plus, and leading
zero tolerance is preserved. Regression coverage added at both the
library level (tests/lib/github-coordination-gh-api.test.js) and the CLI
argument parser (tests/scripts/github-coordination.test.js).

Fixes #3514
2026-10-12 01:25:09 +08:00
haelyra 0d4fd902ca fix: carry verified storage capability and deterministic Canvas checks 2026-10-11 12:49:24 -04:00
haelyra 33372e9970 test: make presence expiry controls deterministic 2026-10-11 12:45:43 -04:00
akushonkamen cbca1dc2cc fix(claude-plugin): target explicitly disabled destinations in scope migration
Issue #3492: scope migration aborted with generic verification errors in
two real states. When the destination scope is installed but inventory
reports enabled:false because the user explicitly set
enabledPlugins["ecc@ecc"]: false, setup now throws a targeted
DESTINATION_EXPLICITLY_DISABLED error naming the setting and file and
offering the native claude plugin enable recovery, leaving the user's
choice untouched. The generic not-enabled error also gains the enable
command in its recovery list.

uninstallSource now accepts a destination-only state when the source
scope is gone from the fresh inventory even if the uninstall command
errored, and SOURCE_UNINSTALL_FAILED recovery no longer suggests
uninstalling a source scope the inventory no longer lists; it prepends
claude plugin list --json so the stuck state can be inspected.

Tests: node tests/lib/claude-scope-migration.test.js — five new cases
cover explicitly disabled resume and single-scope destinations across
user/project/local settings files, the generic fallback keeping the
enable path, an uninstall error with the source already absent, and
non-circular recovery. 16 -> 21 passing; red against unpatched scripts
is 15 passed / 6 failed. Full-suite failure set is byte-identical to
unpatched HEAD in a zero-dependency environment.

Also sync the pinned SOURCE_UNINSTALL_FAILED recovery in
tests/scripts/setup.test.js ('migration JSON failures retain phase,
scopes, and exact recovery') with the new claude plugin list --json
diagnostic: 29 passed / 1 failed on the patched tree before the sync,
30 passed / 0 failed after, identical to clean 2d515e4f.
2026-10-11 20:48:43 +08:00