Commit Graph
1780 Commits
Author SHA1 Message Date
Santhi Prakash 616716f370 test: isolate cost tracker cache fixtures 2026-08-28 16:23:25 -04:00
Santhi PrakashandPaperclip 5ee14cb2af test(hooks): use unique session IDs in Sonnet 5 pricing tests
Prevent stale /tmp/harness-cost cache files from affecting Sonnet 5, dated,
near-miss, and cache-rate pricing tests by using Date.now() in each session ID.

Co-Authored-By: Paperclip <noreply@paperclip.ing>
2026-08-28 16:23:25 -04:00
Santhi PrakashandPaperclip f5d0b295cb test(hooks): expand Sonnet 5 cost-tracker coverage for cache and model matching
- Add cache write/read token pricing test for Sonnet 5.
- Add dated Sonnet 5 ID and claude-sonnet-50 near-miss regression tests.
- Keep Sonnet 4.6 standard rate distinction intact.

Co-Authored-By: Paperclip <noreply@paperclip.ing>
2026-08-28 16:23:25 -04:00
haelyra 950caaaae1 fix(hooks): preserve short sessions and quote evolved metadata 2026-08-28 16:08:19 -04:00
haelyra b7d6c61b1e Merge remote-tracking branch 'origin/main' into maint/pr-2870-current 2026-08-28 16:03:10 -04:00
haelyra 1953a749a2 Merge remote-tracking branch 'origin/main' into maint/pr-2873-current 2026-08-28 15:59:31 -04:00
Aditya Datta cb9dfabcf0 test: pin Ollama whitespace normalization 2026-08-28 15:59:31 -04:00
Samarjeet Singh Tomar 5caf398a91 refactor(install): expose pure manifest planner 2026-08-27 23:34:48 -05:00
haelyra 51982fdab1 test(opencode): verify recovery through descriptors 2026-08-25 17:27:26 -04:00
haelyra aaaff77ef9 test(opencode): avoid path race in recovery fixture 2026-08-25 17:23:24 -04:00
haelyra 204cc2d2a3 fix(release): stage ECC 2.2 launch safely 2026-08-25 17:19:45 -04:00
haelyra d6d0c4e696 test(nasiko): isolate malformed lock fixture 2026-08-25 13:52:19 -04:00
haelyra e10c4bb5bf fix(nasiko): use descriptor lock identity 2026-08-25 13:49:26 -04:00
haelyra 307bbd53a6 fix(nasiko): harden lifecycle recovery 2026-08-25 13:34:58 -04:00
haelyra d66eaf116f test(opencode): canonicalize Windows path expectations 2026-08-25 12:55:02 -04:00
haelyra 5aa660219e test(opencode): isolate environment regression processes 2026-08-25 12:37:39 -04:00
haelyra 856733263c test(opencode): cover legacy upgrade edge cases 2026-08-25 12:28:23 -04:00
haelyra 2331afbfd3 test(opencode): reproduce ambient config leakage 2026-08-25 12:14:58 -04:00
Nguyen Thanh Dat e3ab97915f test(observer-status): set process.exitCode so the summary always flushes
Review feedback on #2878. stdout is async when it is a pipe, which is exactly
how tests/run-all.js runs these files, and process.exit() does not wait for
pending writes — so exiting that way can drop the Passed:/Failed: lines the
aggregator totals, defeating the previous commit. The sibling
tests/ci/ito-*-skill.test.js files already use process.exitCode.

Still exits 1 on a broken counter (Passed: 4, Failed: 7) and 0 when clean.
2026-08-25 18:37:00 +07:00
Nguyen Thanh Dat 787154775e test(locale-agents): set process.exitCode so the summary always flushes
Review feedback on #2879. stdout is async when it is a pipe, which is exactly
how tests/run-all.js runs these files, and process.exit() does not wait for
pending writes — so exiting that way can drop the Passed:/Failed: lines the
aggregator totals, defeating the point of printing them. The sibling
tests/ci/ito-*-skill.test.js files already use process.exitCode.

Still exits 1 on drift and 0 when clean.
2026-08-25 18:35:26 +07:00
Nguyen Thanh Dat 5f79c39687 test(locale-agents): derive the file lists instead of accumulating them
Review feedback on #2879, citing rules/common/coding-style.md ("Immutability
(CRITICAL): ALWAYS create new objects, NEVER mutate existing ones"). Build one
list of locale entries, then derive the matched and orphaned sets from it with
filter/map, and build the canonical Map the same way.

Behaviour is unchanged and both mutations still fail: reintroducing the drift in
one locale file gives `Passed: 4, Failed: 3`, an orphan gives `Passed: 6,
Failed: 1`, both exit 1.
2026-08-25 18:34:12 +07:00
Nguyen Thanh Dat 4f18d1eb87 test(locale-agents): fail on a locale doc whose agent no longer exists
Review feedback on #2879: a locale file with no counterpart in `agents/` was
skipped, so every other case here silently passed over it. Retiring an agent
would leave its seven translations behind with nothing to compare against and
nothing to report it.

Collect those files while building the list and assert the collection is empty.
Zero today across 199 locale docs, so this changes no current result; planting
one orphan takes the file to `Passed: 6, Failed: 1` and exit 1.
2026-08-25 18:04:38 +07:00
Nguyen Thanh Dat ed6e5a42ee test(observer-status): report every case, not just the first failure
Review feedback on #2878: the static assertions ran at the top level, so a
failure exited the process before the `Passed:`/`Failed:` lines. tests/run-all.js
totals those tokens, so the per-case counts were lost — it still went red via the
non-zero exit, but the granular numbers were not in the totals.

Route every case through the same `runTest()` wrapper the file already used for
the integration cases, and build the case list inside a try so a missing or
renamed `instinct-cli.py` / `start-observer.sh` is a reported failure rather than
a crash.

Reverting the fix now prints `Passed: 4, Failed: 7` and exits 1, naming all seven
broken expectations instead of stopping at the first.
2026-08-25 18:03:35 +07:00
Nguyen Thanh Dat 3afd97b4aa fix(docs): stop translated agent docs from contradicting the shipped agent
`scripts/ci/validate-agents.js` reads only `agents/`, so the translated copies
under `docs/<locale>/agents/` were never validated against the agent they
describe — and drifted.

Two kinds of drift, both machine-checkable and both wrong in the same
direction (the translations were made from an older revision and never
re-synced):

- **Model tier, 39 files.** Every one names a costlier tier than ships:
  haiku -> sonnet, sonnet -> opus. `security-reviewer` reads `opus` in ja-JP
  and zh-TW; it ships `sonnet`.
- **Tool set, 15 files.** `security-reviewer` and `database-reviewer` list
  `Write` and `Edit` in **all seven locales** — both agents ship read-only
  (`Read, Grep, Glob, Bash`). `seo-specialist` in zh-CN adds a `Bash` the
  canonical agent does not have. Documenting a reviewer as able to write is
  the kind of inaccuracy someone auditing what these agents can touch would
  act on.

Sync `model` and the `tools` set to canonical, preserving each locale's
existing list style so the diff is only the values that were wrong. The
`["a", "b"]` vs `a, b` formatting difference is left alone: it is consistent
per locale and carries no meaning.

Add `tests/ci/locale-agent-frontmatter.test.js` so this cannot drift back.
`tools` is compared as a set, not a string, so the style difference stays
legal; prose is not compared at all. The last case pins the specific failure:
a canonical read-only agent may never be documented with Write or Edit.

`.kiro/agents/` is deliberately excluded — it uses a different schema
(`allowedTools: [read, shell]`, no `model`), not a translation of this one.
2026-08-25 17:39:43 +07:00
Nguyen Thanh Dat 9af6606142 fix(continuous-learning-v2): count every instinct extension in observer status (#2859)
`start-observer.sh status` globbed `*.yaml`, but the observer prompt tells the
analyzer to write `${INSTINCTS_DIR}/<id>.md` and the loader accepts
`.yaml`, `.yml`, and `.md` (ALLOWED_INSTINCT_EXTENSIONS in
scripts/instinct-cli.py). The one command an operator runs to confirm that
learning works therefore reported `Instincts: 0` on a healthy install — which
is indistinguishable from a silently dead observer, exactly the failure the
status check exists to surface.

Match the loader instead of one of its three extensions, and match how it
enumerates them: `Path.iterdir()` is top level only and `is_file()` skips
directories, so `-maxdepth 1 -type f`; `suffix.lower()` makes the comparison
case-insensitive, so `-iname`. `tr` drops the column padding BSD `wc` emits,
which is why the reported output read `Instincts:        0`.

Verified end to end against the shipped script with 3 `.md`, one `.yaml`, one
`.yml`, one `.YAML`, a `notes.txt`, and a nested `.md`: 1 before, 6 after —
the same six files the loader picks up.
2026-08-25 17:12:08 +07:00
Aditya Datta 9d233aaa63 Trim provider names in prompt builder 2026-08-24 22:27:42 -03:00
aoright 4d8893f607 fix(ci): add tool cache directories to check-unicode-safety ignore list
Signed-off-by: aoright <102943475+aoright@users.noreply.github.com>
2026-08-24 22:27:39 -03:00
cadenli c5ea82f6bf fix: accept standard tools list metadata 2026-08-24 22:27:31 -03:00
bengio777 06ac5b8a49 fix(hooks): treat HTTP 404 MCP probes as reachable
The mcp-health-check preflight probes HTTP MCP servers with a bare GET.
Some Streamable HTTP servers route only POST /mcp and answer a bare GET
with 404 (Paper Desktop 0.5.3 is one). The probe scored that as down and
blocked every tool call for the server indefinitely, since the 30s
backoff just re-probes and re-fails.

A routed HTTP response of any status proves the endpoint is reachable,
which is all this preflight claims to check -- 400/401/403/405/406 are
already treated this way for the same reason. Add 404 to the set and let
the real MCP client validate the endpoint.

Adds a regression test that stands up a POST-only server (404 on GET,
200 on POST /mcp); it fails on the current code and passes with the fix.
2026-08-24 22:27:30 -03:00
Conor DohertyandClaude Opus 5 66b1aad3f2 fix: probe POST-only Streamable HTTP MCP servers before marking them dead
The preflight probe in mcp-health-check only ever sent a bare GET to the
server URL. Some Streamable HTTP MCP servers route POST exclusively and
answer any GET with 404 — api.telnyx.com/v2/mcp is one — so the probe
failed permanently against a perfectly healthy server.

404 is not in HEALTHY_HTTP_CODES, so every probe failed, the backoff
compounded to the 10-minute ceiling, and the hook blocked every tool call
for that server before it left the machine while `claude mcp list` still
reported it Connected.

Replay a failed GET as a real JSON-RPC initialize POST and accept that as
proof of life. Whitelisting 404 was the alternative, but it would mask
genuine outages on every other server.

Adds a regression test with a POST-only server that 404s all GETs and
validates the initialize body; it fails without this change.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-08-24 22:27:29 -03:00
Amir Fathi 08c1c4073c fix(lib): remove unused cost-estimate.js duplicate rate table
cost-estimate.js carries its own copy of the stale Opus/Haiku/Sonnet rate
table already reported in #2574, but grepping every .js/.json/.md file
outside node_modules turns up zero callers besides its own test. It was
added in 940135e alongside the statusline observability hooks and never
wired into any of them.

The maintainer's comment on #2656 named two acceptable outcomes: remove
the unused duplicate, or share one rate source with the live tracker.
cost-tracker.js's own fix (#2574) has not landed yet, so sharing its
table now would import numbers that are still wrong. Removing the dead
file is the smaller, immediately-correct step.

Fixes #2656
2026-08-24 22:26:37 -03:00
Suliman Abdulrazzaq ef68f816d1 fix(gan): grant evaluator Playwright tools 2026-08-24 22:26:25 -03:00
Suliman Abdulrazzaq b7faf3d70e fix: pass observer analysis path explicitly 2026-08-24 22:26:24 -03:00
John EllisonandClaude Opus 5 7aa071c5e9 fix(continuous-learning-v2): emit loadable frontmatter from evolve --generate
Artifacts written by `evolve --generate` are inert: Claude Code (and every
spec-compliant Agent Skills client) injects only `name` + `description` at
startup and will not load an artifact missing them.

Today the generator writes:
  - skills:   `# {name}` with no frontmatter block at all
  - commands: `# {cmd_name}` with no frontmatter block at all
  - agents:   `model`/`tools` only, no `name`, no `description`

So the whole evolve pipeline terminates in files that can never load. I hit
this on a real install: 12 generated artifacts across two projects, none of
which Claude Code had ever seen.

This adds a `_evolved_description()` helper and emits proper frontmatter for
all three artifact kinds. The description is sanitised for the two things that
break loaders: `: ` in an unquoted scalar (rejected by strict YAML parsers)
and `<`/`>` (system-prompt injection risk).

Adds two tests to tests/scripts/instinct-cli-evolve-generate.test.js. Both
fail against current main and pass with this change.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-08-24 22:26:22 -03:00
Matt Van Horn 40c8235d48 fix: address self-review findings 2026-08-24 22:26:22 -03:00
haelyra dac154eff6 test(opencode): cover override lifecycle routing 2026-08-24 21:18:54 -04:00
haelyra 01779a4a2b test(release): cover final review blockers 2026-08-24 21:14:29 -04:00
haelyra 7d9f70c501 test(release): enforce release-note filename convention 2026-08-24 21:08:51 -04:00
haelyra e3f2a537f9 test(opencode): cover legacy migration boundaries 2026-08-24 21:01:06 -04:00
haelyra e3a1ac6f3f fix(opencode): migrate legacy managed home installs 2026-08-24 20:59:57 -04:00
haelyra 47d629633b test(release): require packed uninstall skill cleanup 2026-08-24 20:56:27 -04:00
haelyra 55a2d4823b test(opencode): cover legacy managed root migration 2026-08-24 20:53:53 -04:00
haelyra 17ab179ecc test(release): enforce versioned notes contract 2026-08-24 20:40:35 -04:00
haelyra a504b19411 test(release): derive reviewed notes from tag 2026-08-24 20:34:26 -04:00
haelyra 8b5ef235ff test(release): accept install command result shape 2026-08-24 20:23:33 -04:00
haelyra d0e14ed8f6 test(opencode): align repair fixtures with canonical home 2026-08-24 20:17:03 -04:00
haelyra 2c5a91a1d6 fix(release): make ECC 2.2 ready to publish 2026-08-24 20:15:12 -04:00
haelyra 528dbea019 test(security): reject symlinked guided install sources 2026-08-24 20:13:18 -04:00
dajiaohuang da1faf1400 test(pack): select the requested package 2026-08-24 20:09:12 -04:00
dajiaohuang 6d42f32ca8 fix(tests): support npm pack object output 2026-08-24 20:09:12 -04:00