Addresses Greptile review on PR #407:
- P1: setup-keychain.sh ALL_KEYS was missing GOOGLE_GENAI_API_KEY and
XIAOHONGSHU_API_BASE relative to _load_keychain's inline list, so users
manually storing those keys would not see them in --list and the
interactive prompt would never offer to set them.
Hoist the canonical key list into lib/env.py::KEYCHAIN_KEYS, have
get_config() pass it through, and add a parity test that parses
ALL_KEYS out of setup-keychain.sh and asserts equality. Drift is now
caught at CI time instead of after a user reports a missing key.
- P2: os.environ.get("USER", "") silently returned "" under sudo, in
Docker without --env USER, or in CI runners that strip USER. The
resulting `security find-generic-password -a ""` call would never
match items stored by setup-keychain.sh, so all lookups silently
returned nothing. Fall back to pwd.getpwuid(os.getuid()).pw_name when
USER is absent.
The P2 process-listing comment ("secret visible briefly via ps because
security has no stdin path for -w") has no clean fix — the README
already documents the manual `security add-generic-password` invocation
as an alternative for users with strict secret hygiene.
Adds the macOS Keychain as the lowest-priority credential source on Darwin.
Items stored as generic passwords with service name "last30days-<KEY>" for
the current user are picked up automatically by get_config() — file env
and process env still win on collision.
No new config knob: behavior is strictly additive. On non-Darwin (or when
the `security` binary is missing) the loader is a no-op, so Linux/Windows
behavior is unchanged.
Priority (highest wins):
1. Environment variables
2. .claude/last30days.env (per-project)
3. ~/.config/last30days/.env (global)
4. macOS Keychain items prefixed last30days- (new)
Includes:
- lib/env.py: KEYCHAIN_SERVICE_PREFIX constant, _load_keychain helper
(platform-gated, shutil.which-gated, subprocess-error tolerant),
wiring into get_config before get_openai_auth so OPENAI_API_KEY can
come from Keychain too, _CONFIG_SOURCE reports "keychain" when no
file source is present.
- scripts/setup-keychain.sh: bash helper with interactive set,
--list, --delete, --replace modes. Uses `security add-generic-password`.
- tests/test_env_keychain.py: 12 tests covering platform gate,
missing-binary gate, success path, whitespace stripping, subprocess
errors swallowed, get_config precedence, and an OPENAI_AUTH wiring
regression test.
- tests/test_env_cookies.py: existing integration test mocks the new
_load_keychain hook so it stays hermetic on Darwin developer
machines that have real keychain entries.
- README.md: new "macOS Keychain (optional)" subsection under
"Bring your own keys" documenting setup-keychain.sh and the manual
`security add-generic-password` invocation.
Tested on macOS with a populated keychain and against the existing pytest
suite — CI-tracked tests (test_plugin_contract.py, test_version_consistency.py)
plus all env-touching tests pass. Pre-existing unrelated failures in
test_store.py / test_watchlist_commands.py / test_setup_openclaw.py /
test_footer_nudge_suppression.py are untouched.
Every job sync.sh did has a better replacement:
- Per-harness skill dirs (~/.claude/skills, ~/.codex/skills, ~/.agents/skills):
`npx skills add . -g -y` writes to every detected harness's home dir and
uses symlinks by default. Edits propagate live — no re-deploy step.
- Hermes (~/.hermes/skills/research/last30days):
`hermes skills install mvanhorn/last30days-skill --force` pulls from
GitHub and handles the deploy itself. The script wrapping was redundant.
- OpenClaw variant: `clawhub install last30days-official` is what users
already run per the README; the maintainer doesn't need a separate
variant-deploy step in the public repo's scripts.
- Claude marketplace cache (~/.claude/plugins/cache/...): this was a
"test against the official install path" hack we shouldn't have been
recommending. With PR #400's resolver collapse, STEP 0 no longer
enforces the cache as the only valid SKILL.md location. Just install
the skill normally via `npx skills` or the marketplace.
Cleanup:
- DELETE skills/last30days/scripts/sync.sh
- tests/test_version_consistency.py — drop test_sync_cache_path_uses_skill_version
- CLAUDE.md — replace the sync.sh command + rule with `npx skills add . -g -y`
- HERMES_SETUP.md — Installation now uses `hermes skills install --force`;
developer-alternative section shows the symlink pattern for live editing
- render.py — _skill_version docstring no longer attributes the
".claude-plugin absent" case to sync.sh; explains it via per-harness
install paths in general
- .github/PULL_REQUEST_TEMPLATE.md — drop the "Ran bash scripts/sync.sh"
checklist item
CHANGELOG and historical docs (release notes, plan files) keep their
existing sync.sh mentions as accurate history.
Greptile's review flagged mktemp -t as non-portable between BSD and GNU.
The suggested replacement (mktemp "$TMPDIR/...XXXXXX.json") is correct
about dropping -t but still puts X's in the middle of the template name
(XXXXXX.json), which BSD mktemp does not substitute — only X's at the
end of the basename are replaced on BSD. Verified on macOS:
mktemp "$TMPDIR/last30days-test.XXXXXX.json"
→ /var/folders/.../last30days-test.XXXXXX.json (X's left literal)
The fully portable form uses trailing X's and drops the .json suffix
(engine reads by path, not extension):
mktemp "$TMPDIR/last30days-test.XXXXXX"
→ /var/folders/.../last30days-test.DXAHzR (X's substituted)
Verified on bash and zsh, BSD/macOS. GNU/Linux is already fine since
GNU substitutes X's wherever they appear in the basename.
Applied to both --competitors-plan (comparison-mode block) and --plan
(Step 1 block) tmpfile writes.
Closes#403.
The SKILL.md templates instructed the model to invoke the engine with
inline single-quoted JSON: `--plan '$JSON'` and `--competitors-plan '{...}'`.
When any resolved field value contained an apostrophe (common in `context`
strings like "McDonald's", "people's choice", or contracted forms like
"don't", "won't"), the inner `'` closed the outer single-quote and broke
shell parsing before the engine was even invoked.
Observed during PR #400 testing: a Codex run hit the trap and self-healed
by re-encoding, wasting one engine invocation and ~30s of latency.
Fix: switch both templates to the heredoc + tmpfile pattern. The engine's
`parse_plan()` and `parse_competitors_plan()` already check
`os.path.isfile(plan_str)` and read from disk — only the SKILL.md prose
needed to change.
The quoted heredoc marker (<<'PLAN_EOF') is load-bearing: it suppresses
shell interpolation so apostrophes, $, backticks, etc. pass through verbatim.
A trap on EXIT cleans up the tmpfile after the engine call returns.
LAW 7's "MUST contain --plan" self-check guidance and Step 1's invocation
example both updated to reference the file form. Comparison-mode invocation
block updated the same way for --competitors-plan.
Version bump 3.2.2 -> 3.2.3 because this is a behavior change users
running comparison-mode queries will notice (no more "shell quoting error,
retrying" sequences on apostrophe-containing context strings).
Two real bugs flagged in the automated review of PR #400; both small.
1. render.py::_skill_version manifest with no "version" key
`json.loads(manifest.read_text()).get("version", "?")` returned "?"
immediately on a valid JSON manifest that lacked the "version" key,
never falling through to the SKILL.md frontmatter fallback. Contradicted
the docstring's "Returns '?' only if both sources are missing" contract.
Same shape if version is present but empty string ("" produces the
broken badge `🌐 last30days v · synced ...`).
Fix: pull the version out of the parsed dict, then `continue` to the
next ancestor if it's None or empty. Falls through to the SKILL.md
walk only after exhausting every ancestor.
2. SKILL.md STEP 0 re-read target hardcoded to nested cache layout
STEP 0 told the model to re-read from
`$CLAUDE_CACHE_LATEST/skills/last30days/SKILL.md` — the new nested
layout. But Step 1's resolver explicitly handles both shapes
(nested `{cache}/{version}/skills/last30days/` and flat
`{cache}/{version}/`), noting "Both shapes ship in the wild." On an
install where the highest-versioned cache happens to be the older flat
shape, STEP 0's re-read target wouldn't exist; the model would silently
stay on the stale marketplaces/ copy STEP 0 was supposed to move it
away from — the exact failure mode this guard was added to prevent.
Fix: extend the STEP 0 bash to resolve $CLAUDE_CACHE_SKILL_MD by
probing both layouts, then have the model hop to that resolved path
instead of constructing the path from a hardcoded suffix.
Two new tests in tests/test_skill_version.py cover the missing-key and
empty-string cases for fix 1. Fix 2 is exercised via the bash probe at
verify time (the STEP 0 prose-contract test isn't unit-testable from
Python, but the dual-layout bash is verified to resolve to the correct
SKILL.md on both shapes).
Stale finding skipped: greptile also flagged a missing try/except on the
SKILL.md read_text() call, but that was already addressed during the
ce-code-review safe_auto pass earlier in this PR — current code wraps it
in `try/except (OSError, UnicodeDecodeError)`, strictly more defensive
than the suggested fix.
zsh errors on globs that match nothing instead of returning the literal
pattern (bash's default), and `2>/dev/null` does not suppress the error
because it comes from the shell's glob expansion before `ls` even runs.
Under Codex (which executes the SKILL.md bash via zsh), STEP 0 and the
Step 1 / comparison-mode resolvers emitted noisy "no matches found"
errors on machines without a Claude plugin cache populated.
Replaces all three `ls -d $HOME/.claude/plugins/cache/last30days-skill/last30days/*/`
invocations with `find ... -mindepth 1 -maxdepth 1 -type d 2>/dev/null`.
find is POSIX-portable, errors silently when the base dir doesn't exist,
and never triggers shell glob errors. `sort -V | tail -1` precedence
preserved (verified: picks 3.10.0 over 3.2.1 over 3.1.0). Trailing-slash
strip removed because find doesn't append slashes.
Observed in Codex session running /last30days against PR #400 with the
Claude plugin cache deleted - bash output was:
zsh:1: no matches found: /Users/.../last30days/*/
After fix: clean empty output, exit 0, STEP 0 correctly treats it as
"no cache present, do not hop", resolver falls through to per-harness
skill dirs as designed.
12 fixes from the multi-agent code review on PR #400:
Version 3.2.1 -> 3.2.2 across all manifests (SKILL.md frontmatter + body
header, pyproject.toml, .claude-plugin/{plugin,marketplace}.json, sync.sh
cache path). The PR ships observable behavior changes (STEP 0 logic flip,
resolver order change, badge fallback) that should not silently appear
under the same version number — the new fallback reads SKILL.md version
directly so the badge would otherwise be misleading.
render.py::_skill_version:
- `import re` moved to module top
- _VERSION_RE extracted as a module-level compiled pattern that accepts
double-quoted, single-quoted, and unquoted YAML version scalars
- `break` -> `continue` on corrupt manifest, so a corrupt inner manifest
no longer shadows a valid outer one
- Wrap SKILL.md read_text() in try/except for UnicodeDecodeError to keep
badge emission from crashing on mis-encoded SKILL.md
- Docstring clarifies precedence; inline comment marks the fallback boundary
between the manifest walk and the SKILL.md walk
tests/test_skill_version.py (new): 7 unit tests for the fallback paths
(manifest absent, manifest corrupt, corrupt-inner + valid-outer, both
absent, SKILL.md without version, single-quoted, unquoted).
tests/test_plugin_contract.py: tombstone test asserting .codex-plugin/
stays removed (was the only CI guard against accidental reintroduction).
SKILL.md:
- STEP 0 bash echoes CLAUDE_CACHE_LATEST so the model can see the
resolved value when deciding whether to hop
- "Both shapes ship in the wild" comment now names the two cache layouts
(nested {cache}/{version}/skills/last30days/ vs flat {cache}/{version}/)
- Comparison-mode bash invocation gets its own inline SKILL_ROOT resolver
(latent gap: the contract tells the model to skip Step 1 on comparison
queries, so SKILL_ROOT was previously unset there)
CHANGELOG.md: [Unreleased] entries for the resolver rewrite and the
breaking removal of Codex native-plugin support.
All 9 reviewer personas surfaced findings; 3 cross-reviewer corroboration
clusters were promoted (import re, "both shapes" comment, missing fallback
tests). Maintainability follow-up flagged: regex now duplicated across
render.py and 2 test files; could consolidate via shared lib/skill_meta.py
helper in a future PR.
STEP 0 (CANONICAL PATH SELF-CHECK) used to force any SKILL.md load that wasn't
under $HOME/.claude/plugins/cache/last30days-skill/last30days/{version}/ to
re-Read from there. That guard is Claude-Code-specific (defends against the
marketplaces/ stale-clone bug) and broke under non-Claude installers like
`npx skills add`, ~/.codex/skills/, and ~/.agents/skills/.
The new STEP 0 narrows the check to its actual target: fire only when the
loaded SKILL.md path contains /.claude/plugins/marketplaces/. Every other
install path is trusted. The 2026-04-22 incident workaround is preserved
without breaking other harnesses.
Step 1 SKILL_ROOT resolver collapses the Codex-first / Claude-fallback /
CWD-fallback chain into a single precedence walk: Claude plugin cache
(versioned) first, then ~/.codex/skills, ~/.agents/skills, repo checkout,
./.skills/last30days (npx skills install dir), CWD, and GEMINI_EXTENSION_DIR.
Also drops Codex native plugin support: .codex-plugin/plugin.json is deleted,
the badge VERSION jq fallback in line 108 stops looking at it, and render.py's
_skill_version no longer scans for it. Codex users install via `npx skills add`
or the per-harness skill dir going forward.
render.py::_skill_version gains a SKILL.md frontmatter fallback so the badge
no longer emits `v?` on install dirs that sync.sh populates (which don't
include .claude-plugin/plugin.json).
sync.sh was written against the layout of mvanhorn/last30days-skill-private
(`.../cache/last30days-skill-private/last30days-3/{version}`) and that path
was never updated when this public repo got its own copy. Running sync.sh
from here populated the BETA channel's cache (`/last30days-beta`) instead
of this repo's own `/last30days` cache, so devs working in this repo could
not test their changes via the public slash command without waiting for a
marketplace release.
Path now derives from this repo's own manifests:
- marketplace name `last30days-skill` (.claude-plugin/marketplace.json)
- plugin name `last30days` (.claude-plugin/plugin.json)
Drops the `last30days-3-nogem` target along with it - that's a private-repo
variant with no public equivalent.
Updates test_sync_cache_path_uses_skill_version to assert the new path
pattern and clarifies the COMMON_TARGETS comment so the next person editing
it understands which marketplace/plugin name segments come from where.
Five provider modules (pinterest, threads, instagram, tiktok, youtube_yt)
and watchlist.py each carried a try/except `requests` import with parallel
urllib + requests branches. The urllib path already used the
stdlib-only wrapper at `lib/http.py` (retries, 429 handling, HTTPError).
This collapses every dual-branch into a single `http.get`/`http.post`
call and removes the `requests` dependency from `pyproject.toml`.
Also drops 4 transitive deps (urllib3, certifi, charset-normalizer, idna)
from the lockfile, leaving the skill stdlib-only at runtime.
Tests for tiktok comments and watchlist delivery were rewritten to mock
`lib.http` directly instead of the now-removed `requests` module.
Out of scope but flagged during review: the 13 surviving SC call sites
share a near-identical scaffold and would benefit from a
`http.scrapecreators_get(url, params, token, ...)` helper. Filed for a
follow-up PR rather than expanding scope here.
test_sync_cache_path_uses_skill_version asserts that sync.sh's plugin
cache path includes the version from SKILL.md frontmatter. The frontmatter
moved to 3.2.1 in #371 but sync.sh still pointed at 3.2.0, leaving CI red
on every PR.
Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com>
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* feat(digg): bump POSTS_PER_CLUSTER to 5 and render limit to 3
Match the per-item enrichment cap and inline-display cap used by the
other sources (Reddit, HN, YouTube, TikTok, GitHub all use 5 fetched /
3 displayed). At the previous 3/2 caps the engine routinely truncated
cluster context — a recent run on cli-printing-press lost the Jason
Calacanis quote tweet entirely because the display cut off after Garry
Tan's first two posts.
* feat(digg): rename 'Digg AI 1000' to 'Digg' in user-facing strings
Drop the 'AI 1000' suffix from the footer line, source label, inline
quote attribution ('via Digg'), why_relevant, container, mock title,
SKILL.md source list, and README sources table. Internal code comments
and docstrings still reference the upstream Digg AI 1000 product.
Bumps version to 3.2.1 and adds a CHANGELOG entry covering this rename
and the POSTS_PER_CLUSTER / render-limit bumps from the prior commit.
---------
Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com>
* chore(release): v3.2.0
Bumps plugin/marketplace/codex/pyproject versions from 3.1.1 to 3.2.0.
Promotes the Unreleased CHANGELOG entries (--emit=html, Digg AI 1000
source) to the 3.2.0 release section.
* chore(release): bump SKILL.md header and sync.sh path to 3.2.0
---------
Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com>
* feat(digg): add Digg AI 1000 source module with cluster search and post enrichment
- search_digg shells out to digg-pp-cli with --since 30d --agent
- parse_digg_response normalizes clusters to last30days dict shape
- enrich_with_top_posts attaches top-ranked X posts to top-K clusters
- shutil.which gate plus subproc.run_with_timeout discipline matches
bird_x.py / youtube_yt.py patterns
25 unit tests cover parse, age window, relevance, binary-missing
fallback, timeout recovery, and partial enrichment failures.
* feat(digg): wire Digg source into pipeline, normalize, signals, and render
pipeline.py:
- Import digg, add to MOCK_AVAILABLE_SOURCES, gate via shutil.which
- Dispatch case calls search_digg + parse_digg_response, runs
enrich_with_top_posts at default/deep depth
- Mock fixture includes one enriched cluster + one bare cluster
normalize.py:
- _normalize_digg maps cluster dicts to SourceItem with
container='Digg AI 1000' and metadata.posts pass-through
signals.py:
- SOURCE_QUALITY['digg'] = 0.85 (top tier alongside YouTube,
reflecting Digg's curatorial layer)
- ENGAGEMENT_WEIGHTS['digg'] balances postCount, uniqueAuthors,
and the rank_score derived from Digg's curatorial position
render.py:
- SOURCE_LABELS['digg'] = 'Digg AI 1000'
- _FOOTER_SOURCES adds '⛏️ Digg AI 1000' line after GitHub
- ENGAGEMENT_DISPLAY mirrors footer keys
- New _digg_posts_for + _format_digg_quote helpers emit inline
'@handle via Digg AI 1000' quotes for clusters with attached X
posts; both compact and full-dump renderers call them
* feat(digg): polish per-item engagement display and progress label
- ENGAGEMENT_DISPLAY for digg uses 'posts' / 'auth' to match the
codebase abbreviation convention (HN: 'pts'/'cmt', X: 'rt'/'re')
- Footer item word changes from 'story' to 'cluster' to dodge the
pre-existing naive plural in _footer_line_for_source ('storys')
and to match Digg's actual data model
- ui.py SOURCE_COMPLETION_META adds digg with correct 'cluster'/
'clusters' plural so 'Research complete' shows 'Digg: N clusters'
* feat(digg): document Digg AI 1000 source in skill, README, and changelog
- planner.py SOURCE_CAPABILITIES adds digg with discussion/social/link
capabilities so the planner offers it through the standard fanout
- SKILL.md ACTIVE_SOURCES_LIST gate includes 'which digg-pp-cli' check
and the source list / available-sources line names digg as opt-in
- README.md Sources table adds the Digg AI 1000 row with the activation
gate so first-time readers see what they get
- CHANGELOG.md Unreleased section calls out the source addition
* fix(digg): enrich post-dedupe so brief survivors carry inline quotes
Pipeline dispatch was attaching X posts to the top-3 items returned by
search, but dedupe later picked different survivors when multiple
clusters compared similar (common for trending topics). The brief
ended up showing clusters with no posts attached even though
enrichment ran successfully on positions 0-2.
Move enrichment to _finalize_items_by_source. The new
digg.enrich_source_items helper reads metadata['clusterUrlId'] and
writes metadata['posts'] in place on the SourceItems that actually
survive dedupe.
Verified live on 'openclaw': 2 surviving clusters, both now carry
real X-post quotes from @sama and @jeremyphoward attributed
'via Digg AI 1000'.
Adds 3 unit tests covering survivor enrichment, non-digg skip, and
clusterUrlId fallback to item_id.
* test(digg): relax live off-topic test to check shape, not emptiness
Digg's live search uses fuzzy/popularity fallback, so an impossible
token can still return some loosely-related clusters. The contract
the pipeline depends on is shape (results is always a list);
token-overlap relevance handles the noise downstream.
---------
Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com>
Adds a one-command shareable HTML mode to /last30days. The skill detects
HTML intent (explicit --emit=html / --emit:html / --html flag in
$ARGUMENTS, or natural-language asks like "give me a shareable brief",
"for Slack", "export as HTML"), runs the normal research + chat synthesis
flow, then saves a self-contained HTML file to
~/Documents/Last30Days/{topic}-brief.html. The synthesis appears in chat
as usual; the HTML is an additional artifact for sharing.
User experience:
/last30days OpenClaw --emit=html
/last30days OpenClaw, give me an HTML brief for Slack
Synthesis prints to chat. Last line of the response: "📎 Shareable brief
saved to ~/Documents/Last30Days/openclaw-brief.html". Open it, drag it
into a message, browser-print to PDF, email it.
Architecture:
- SKILL.md gets a small detection block (triggers + early exit +
MUST/MUST NOT rules + rationale) that points to a reference file.
- references/save-html-brief.md owns the implementation: capture the
synthesis verbatim into a temp file via heredoc, invoke the engine
with --emit=html --synthesis-file, save to disk, append the
confirmation line to chat.
- lib/render.py exposes render_for_html(report, synthesis_md=None) and
render_for_html_comparison(...) -- clean markdown for HTML
conversion. Omits debug file header, model-facing safety note, and
data quality warnings (those stay in engine stderr; recipients can't
act on them in a shared artifact).
- lib/html_render.py is a new module: ~200-line CSS template (dark
mode default, prefers-color-scheme switch, print stylesheet, mobile
breakpoint), stdlib-regex markdown-to-HTML converter, marker-based
META + engine-footer wrapping, PROSE_LABELS registry promoting plain
-text labels to <h2>, colophon builder.
- last30days.py adds --emit=html argparse choice and --synthesis-file
PATH flag (engine still callable directly without the skill in the
loop).
Design:
- Voice-led research brief, not corporate report. Inter + JetBrains
Mono via Google Fonts with full system fallbacks (no FOIT, works
offline). Brand purple #a855f7 (#7c3aed in light mode). Type ramp:
body 17px/400/muted, bold lead-in 17px/600/fg, h2 + .prose-label
20px/600/fg, monospace badge/meta/footer/colophon at 13-13.5px.
- 720px max-width, generous whitespace, no card layouts or shadows.
- Print stylesheet: light theme, A4 margins, [href]::after URL
footnotes, page-break-inside:avoid on the engine footer.
Templated (locked) shell:
- HTML5 boilerplate, Google Fonts <link> with preconnect, all CSS
inline.
- .badge / .meta / .engine-footer / .colophon containers.
Flexible (role-based):
- <h2> rendering covers BOTH plain ## headers (comparison mode per
LAW 4 exception) AND promoted prose labels via PROSE_LABELS
registry. Adding a new SKILL.md prose label is a one-line tuple
addition; no CSS or template changes.
- Marker-based engine boundaries (<!-- META: ... -->,
<!-- PASS-THROUGH FOOTER -->) survive the markdown converter and
get promoted post-conversion. Robust to engine output format
changes.
- Generic markdown-to-HTML for body content; future SKILL.md additions
(new sections, tables, blockquotes) render correctly without code
changes.
Tests: 30 new tests in tests/test_html_render.py covering snapshots
(rich/thin/comparison), CLI parsing, --synthesis-file end-to-end, prose
label promotion, warning exclusion from artifact, parseability via
html.parser, no-script self-containment.
No SKILL.md voice contract changes, no LAWs 1-8 changes, no new pip
dependencies, no JavaScript anywhere.
is_available() only caught FileNotFoundError and TimeoutExpired. On WSL,
a /mnt/c/.../WindowsApps entry on $PATH returns EACCES during exec, and
Python raises PermissionError. That escaped is_available() and crashed
pipeline.diagnose() before any source ran.
Catch OSError instead. It covers FileNotFoundError, PermissionError, and
any other spawn-time OS error, so a non-executable xurl on PATH falls
through to the next backend instead of aborting the run.
bird_x.py and youtube_yt.py had four near-identical copies of the same
subprocess cleanup dance (Popen + os.setsid + communicate(timeout) +
SIGTERM via killpg + proc.kill() fallback + wait(5)). Extract to
lib.subproc.run_with_timeout(), which:
- runs the child in its own process group via os.setsid where available
- raises SubprocTimeout on timeout
- on timeout: SIGTERM the group, fall back to proc.kill(), wait up to 5s
- accepts an on_pid callback so bird_x can still register child PIDs
with last30days.register_child_pid for whole-process cleanup
- captures stdout/stderr as strings in a SubprocResult dataclass
Migrated call sites: _run_bird_search, search_handles inner worker,
search_youtube, fetch_transcript. With the helper in place, the signal
and subprocess imports became dead in both files (plus os in
youtube_yt) and went with them.
Tests: 9 new subproc tests cover success, non-zero exit, stderr capture,
timeout-raises, timeout-kills-group, missing-command, env passthrough,
PID callback, and callback-exception suppression. test_env_v3 and
test_youtube_yt patch subproc.run_with_timeout instead of the removed
bird_x.subprocess and yt-dlp subprocess.
1. N+1 queries in store.store_findings()
The old loop ran one SELECT per finding to check existence, then one
INSERT or UPDATE. 100 findings cost 200 serial SQLite roundtrips.
Now: one batch SELECT with WHERE source_url IN (...) builds a lookup
dict, then executemany() handles all inserts and updates. Query count
stays constant regardless of batch size. Benchmark on 500 findings:
~30ms to ~20ms; gap widens on slower storage.
2. O(n^2) source_items dedup in fusion.weighted_rrf()
Merging an item into an existing candidate ran any(existing.source ==
... for existing in candidate.source_items), linearly scanning a list
that grew with each merge. At 40 candidates with 20 source_items each,
fusion went quadratic. Now tracks (source, item_id) tuples in a
per-candidate set for O(1) lookup. The source_items list itself is
unchanged since other code iterates it.
3. Dead code removal
- providers.GeminiClient.ground_search() and .url_context_json(): zero
callers. Deleted.
- render._top_comment_excerpt(): zero callers. Deleted.
- env.is_reddit_available(): one-line wrapper around get_reddit_source.
Callers can check get_reddit_source(config) is not None directly.
Scoring hot path (_normalize_score_dedupe) re-tokenized the same
ranking_query ~240x per stream: once per item for local_relevance,
plus ~5x per item across snippet windows. Query tokens are immutable
within a stream, so compute them once as relevance.PreparedQuery and
thread through signals.annotate_stream and snippet.extract_best_snippet.
dedupe._PreparedText called normalize_text twice: once in __init__ and
again via get_ngrams. Factor out _ngrams_of_normalized so the prepared
path skips the redundant pass while get_ngrams keeps its public contract.
Behavior unchanged.
_normalize_reddit, _normalize_hackernews, and _normalize_github inlined
the same 5-line comprehension to stringify and space-join the first 3
top_comments' excerpt field. Extract one helper, call it from all three.
The comment field name varies per source (Reddit/GitHub use 'excerpt',
HN uses 'text'), so it's passed as a parameter. Behavior unchanged.
Three independent Opus 4.7 self-debugs on 2026-04-18 converged on the same
root cause of the v3.0.6/v3.0.7 canonical-compliance regression: SKILL.md is
42,860 tokens / 1,478 lines, LAWs lived at line 1094+, every realistic reading
strategy failed to reach them before synthesis.
Unit 1 - Moved the BADGE MANDATORY block and VOICE CONTRACT LAW 1-5 (plus
the formatting-authority preface) from line ~1090 to line ~75 (right after
the SKILL CONTRACT preface, before HOW TO INVOKE THIS SKILL). Every reading
strategy now lands the LAWs in active context before synthesis.
Unit 2 - Engine now emits the badge as the first line of --emit=compact
stdout. Passing through the script output becomes the default-correct
behavior; emitting the badge no longer depends on model compliance. Reads
version from .claude-plugin/plugin.json at runtime with graceful fallback.
Unit 3 - Deleted skills/last30days/SKILL.md stub (231-line v3-spec file).
This was the wrong-file-capture hazard Ron Conway's self-debug identified:
model grabbed the first SKILL.md find surfaced and treated it as
authoritative. Only ONE SKILL.md in the plugin package now.
Diagnoses verbatim:
- Kanye thread: "I read lines 1-600 in chunks, jumped to 300-899, then
stopped. File is 1478 lines. I never saw past ~900."
- Peter thread: "I tried Read once, hit the 25K token cap on a 42,860-token
file, and bailed instead of chunked-reading with offset/limit. I never
opened SKILL.md at all."
- Ron Conway thread: "I read one SKILL.md (231 lines)... the v3 spec stub.
I never opened the operational SKILL.md sitting next to the script."
Validation: direct engine invocation confirms badge at line 1 of compact
output. Module imports clean.
Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com>
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Atomic bump across all four manifests:
- SKILL.md (root)
- skills/last30days/SKILL.md (internal spec)
- .claude-plugin/plugin.json
- gemini-extension.json
CHANGELOG entry documents the skill-upload packaging fix, vendor/ removal,
legacy plans/ removal, and the new scripts/build-skill.sh builder.
Two SKILL.md files declared `name: last30days` with `user-invocable: true`,
which caused strict marketplace validators to reject the plugin with "Some
plugins in this marketplace have validation errors":
- ./SKILL.md (canonical, also reachable via skills/last30days-nux/ symlink)
- ./skills/last30days/SKILL.md (v3 architecture spec, real file)
In v2.9.6, skills/last30days/SKILL.md was a symlink to ../../SKILL.md so
only one skill existed. Commit 0a9ff16 (v3.0.0) added a new real file at
skills/last30days-v3/SKILL.md, and commit 9be0780 then renamed that
directory to skills/last30days/, replacing the original symlink with a
different real file. The collision has been live since v3.0.0 shipped.
This change:
- Renames skills/last30days/SKILL.md to name: last30days-v3-spec and sets
user-invocable: false. The file stays in place as internal architecture
documentation, but it no longer competes with the canonical skill.
- Fixes README.md link that pointed to the deleted skills/last30days-v3/
path (left over from the rename).
- Removes a stale variants/open/SKILL.md reference (variants/open was
deleted in v3.0.0).
After the change, only one canonical name=last30days user-invocable=true
skill exists (the root SKILL.md, also reachable via the
skills/last30days-nux/ symlink, same inode).
Closes#204.
This contribution was developed with AI assistance (Codex).
Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com>
Add gemini-extension.json manifest with correct array-format settings,
symlink skills/last30days/SKILL.md to root SKILL.md for Gemini skill
discovery, add Gemini install paths to bash for-loop in both main and
open variant, and add Gemini CLI install instructions to README.
Incorporates the good parts of PR #53 (manifest, paths, README) while
avoiding duplicate SKILL.md, tool name scattering, and allowed-tools
pollution that would have created maintenance issues.
Closes#45
Co-Authored-By: Alex Ferrari <alex@thealexferrari.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Trevin Chow