Review showed the "Setting the narrative?" verdict compared across
abstraction levels: a homepage tagline is deliberately broad
("financial infrastructure" covers a chargebacks thread), so
tagline-vs-thread alignment verdicts are unfalsifiable and carry no
information. The signal now ships as PROSE in the entity's narrative
section, fires only when the month's evidence directly bears on the
pitch (supports a specific claim, cuts against one, or is squarely
about the pitched ground), and stays SILENT when the pulse is
orthogonal - omission over a manufactured connection. Claims are
tested at matched altitude (specific claim vs specific thread) and
stay windowed (no trend verbs one 30-day window can't support). The
positioning fetch step survives unchanged and now also grounds the
"What it is" row and brand-noise rejection. All scope gating (people
never, ownerless topics excluded, no pitch from memory) carries over.
The lens criterion is now explicit: an identifiable first party with a
fetchable public pitch - companies, products, services. People NEVER
qualify, even founders whose companies would (MrBeast the company can
get the lens; Jimmy Donaldson the person cannot; "Garry Tan vs Sam
Altman" gets N/A across the row). Ownerless topics (Bitcoin) fail the
same test - a foundation or fan site is not an authoritative first
party. Comparison-row verdicts are also artifact-gated to match the
single-entity beat: a verdict requires positioning fetched THIS run;
if item 6 couldn't run (no WebSearch), the cell is Unclear - the pitch
is never supplied from memory.
The entity-grounding demotion required the full multi-word primary
entity as a contiguous substring, so on-entity items missing a trailing
search descriptor were buried: a 323-pt HN thread "Stripe is friendly
to 'friendly fraud'" scored 0 on a "Stripe payments" query. New
_entity_grounded helper keys on the brand head token; items that never
name the brand still miss it and stay demoted. reddit_keyless
_slot_priority, which had re-implemented the old check while claiming
to mirror rerank's signal, now calls the shared helper so the two
paths cannot diverge.
Comparison tables gain a narrative axis judging whether each entity's
community conversation is about what the entity pitches or about
something else (pricing, rivals, an incident). Backed by a new
mandatory Step 0.55 research item that fetches each entity's CURRENT
first-party positioning (RESOLVED_POSITIONING) instead of relying on
memory, and a narrative-check synthesis beat for single-entity company
runs. The mismatch is the signal: companies usually don't control
their own conversation.
* fix(reddit): relevance-aware comment-enrichment slot selection in keyless path
* docs(changelog): record relevance-aware enrichment fix under Unreleased
---------
Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com>
* test(reddit): add live RSS + shreddit comment fixtures
Captured from reddit.com on 2026-05-29 (search.rss listing + the
/svc/shreddit/comments partial), trimmed to a representative subset plus
two synthetic edge cases (deleted author, negative score) for offline
parser tests.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* feat(http): add keyless get_text helper
Browser-UA text fetch for RSS/HTML endpoints; returns None on any HTTP or
network failure so tiered callers fall through cleanly.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* feat(reddit): keyless RSS discovery (search.rss + listing feeds)
Replaces the now-403 search.json with keyless Atom feeds, normalized to the
existing reddit_public post shape. Scores are placeholder zeros, backfilled
during shreddit enrichment.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* feat(reddit): keyless shreddit comment scraper
Parses <shreddit-comment> elements from /svc/shreddit/comments/r/{sub}/t3_{id}
(score/author/created/permalink + thingId-anchored body) into top comments,
matching reddit_enrich output. Replaces the dead {thread}.json enrichment.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* feat(reddit): tiered keyless orchestrator
Tier 0 one-shot .json (residential bonus) -> Tier 1 RSS discovery ->
Tier 2 shreddit enrichment. Returns [] never raises, so the SC backup
still engages when every keyless tier is empty.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* fix(reddit): route free path through keyless pipeline (.json is dead)
search_reddit_public is now a thin shim over reddit_keyless, so pipeline.py
and other callers need no change. Removes the dead .json enrichment helpers;
search/_parse_posts remain as the demoted Tier 0 attempt.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* fix(reddit): request sort=top so true top comments land on page 1
Guarantees the highest-scored comments are captured even on large threads,
independent of Reddit's default comment sort. Local score re-sort remains.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* feat(reddit): recover post upvote scores via keyless listing partials
The shreddit community-more-posts partial server-renders each post's score
and comment count (works for normal users, not IP-gated), unlike RSS or the
comments endpoint. Use it as a scored discovery source and to backfill scores
onto RSS-discovered posts (subreddits derived from results when not provided).
Ranking now uses real upvote score.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* fix(reddit): listings backfill scores only on bare queries, not discovery
Caught running the full pipeline on a bare topic: deriving subreddits from
noisy RSS results and merging their top/hot listings flooded results with
high-upvote off-topic posts. Now derived-subreddit listings are used only to
backfill scores onto keyword-matched RSS posts; listing cards are merged as
discovery only when the caller explicitly provides subreddits (on-topic).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Add a pytest-discovered tests/conftest.py for the last30days scripts path and
remove duplicate per-file sys.path.insert boilerplate from tests.
Normalize affected imports to rely on the shared scripts path and remove the
now-unneeded E402 suppressions.
Greptile review (PR #438) flagged two issues:
1. search_github and enrich_with_comments both call _resolve_token,
so when GITHUB_TOKEN is absent from config and env the gh-CLI
subprocess (with its 5s timeout) fires twice per query.
2. The no-token early-return envelope `{"items": [], "error": "no token"}`
was missing the `context` key that every other failure path includes,
making the envelope shape inconsistent between the no-token and
fetch-failure cases.
Fix 1: add public github.resolve_token(token) wrapping the existing
_resolve_token. Pipeline calls it once before search and enrich, so
both downstream calls receive an already-resolved (or already-None)
token and skip the fallback chain.
Fix 2: thread core/from_date/to_date/count through the no-token
envelope's `context` key, matching the fetch-failure envelope shape.
parse_github_response was already tolerant of the missing key, but
diagnostics callers that read response["context"]["..."] now get a
consistent dict in both error paths.
Reviewer's suggested code patch for issue 1 was a no-op (it kept the
same _resolve_token(token) call inside enrich_with_comments); the
underlying intent — resolve at the boundary — is what this commit
implements.
search_github returned a normalized List[dict] directly while every
other adapter follows search_X -> dict envelope, parse_X_response ->
list[dict]. The github branch in pipeline._retrieve_stream was the
only one that called search_* and returned (result, {}) without a
parse step. This blocked fixture-driven testing: there was no parse
function to feed a synthetic envelope to.
Split into three:
search_github(...) -> Dict[str, Any]
HTTP fetch only. Returns {"items": [raw items], "context": {core,
from_date, to_date, count}}.
parse_github_response(response) -> List[Dict[str, Any]]
Pure function. Normalizes, date-filters, sorts by relevance.
enrich_with_comments(items, depth, token) -> List[Dict[str, Any]]
Public extraction of the old private _enrich_top_items. Resolves
the token via env / gh CLI fallback so callers don't have to.
Pipeline now does the standard 3-call dance:
response = github.search_github(...)
items = github.parse_github_response(response)
items = github.enrich_with_comments(items, depth=depth, token=token)
Keeping enrich_with_comments in parse_github_response would make parse
impure and force every fixture-driven test to either mock HTTP or
skip enrichment. Splitting it out matches the YouTube adapter's
pattern.
Reddit's public JSON endpoint returns 403 to requests carrying the
generic User-Agent and minimal header set urllib defaults to, while
matching curl requests succeed. Switch to a current-Chrome User-Agent
and add Accept-Language / Accept-Encoding / Connection headers so the
fingerprint matches a normal browser. Reddit now serves gzip when
Accept-Encoding includes it, so decompress the body before JSON parse.
Update the user-agent assertion in tests/test_reddit_public.py to match
the new browser-like string. Closes#199.
Co-authored-by: Franco Carballar <francocarballar@gmail.com>
Rebased onto current main where:
- instagram.py uses unified http.get (not _requests fallback); tests now
mock http.get and assert params/timeout kwargs.
- quality_nudge tests use lib.* import path with sys.path setup.
- README/CONFIGURATION.md memory-dir lines say "defaults to" so they pass
test_no_stray_hardcoded_memory_dir_paths.
Six small additive changes that make the skill correctly understand its
configured sources, plus tests + docs.
User-visible benefits
- LAST30DAYS_STORE=1 in .env turns persistence default-on without
remembering --store on every invocation. Mirrors LAST30DAYS_DEBUG /
LAST30DAYS_SKIP_PREFLIGHT convention.
- SCRAPE_CREATORS_API_KEY (with underscore) accepted as alias for the
canonical name. Matches the spelling used in the vendor's own example
code (Adrian Horning's repo); saves the next user the same diagnostic
rabbit hole.
- Bluesky search now hits api.bsky.app (canonical AppView) instead of
public.api.bsky.app (BunnyCDN-blocked public mirror as of 2026-05-04).
BSKY_SEARCH_HOST env var lets users self-rescue future host migrations
without a code release. Pre-fix: silent 0 Bluesky posts on every run.
- App-password format validator emits a one-shot stderr warning when
BSKY_APP_PASSWORD doesn't match xxxx-xxxx-xxxx-xxxx form. Detect-don't-
gate: createSession still accepts main passwords; the warning helps
users identify a hygiene issue without breaking existing setups.
- Instagram retry on multi-token 500. SC's v2 reels endpoint wraps
Google Search and 500's frequently on multi-word queries; a hashtag-
form retry runs once before bubbling up. Documented vendor instability.
- LAST30DAYS_TRANSCRIPT_TIMEOUT env var (default 30s, was hardcoded 15s).
SC's transcript endpoint regularly takes >15s; the old default was
clipping legitimate responses.
- Silent-failure visibility: new bonus_errored field in the quality
nudge fires when SC is configured but Instagram returned 0 items.
Users see "Bonus source silent: Instagram" instead of unexplained
absence.
- YouTube degraded-ratio false-positive fixed. Captions-disabled videos
can never produce a transcript regardless of yt-dlp version; they're
now subtracted from the denominator so a single uploader-disabled
video doesn't false-trigger the "stale yt-dlp" nudge.
- urllib retry path: status_code attribute typo fix. The Instagram
500-retry was dead code on the urllib branch (getattr(e, 'status', ...)
while http.HTTPError exposes status_code).
Docs
- README.md: added /plugin install last30days step after marketplace add
in three places (the install was previously omitted in the docs).
- CONFIGURATION.md: documented LAST30DAYS_STORE env var, added
BSKY_SEARCH_HOST + app-password format section, mentioned
LAST30DAYS_TRANSCRIPT_TIMEOUT in the Instagram source row.
Test plan
- 43 new unit tests across test_bluesky.py, test_instagram_sc.py,
test_quality_nudge.py, test_youtube_yt.py
- 141 total tests passing in target suite
- Verified end-to-end: /last30days "Toronto resale condo market" with
all 11+ sources active stored 35 new + 5 updated findings, all builder-
PR-style accounts absent (organic agent voice in Instagram + TikTok
results)
Backward compatibility
All changes are strictly additive. Optional kwargs default to None.
New env vars are opt-in. Existing CLI flags untouched. Existing callers
of public functions unaffected.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Two findings from Greptile review on PR #302:
1. tests/test_cli_v3.py:302 - The test asserted run_mock.call_args_list[0]
was the main runner's invocation, but fanout.run_competitor_fanout
submits main + competitors to a ThreadPoolExecutor and iterates with
as_completed. With zero-latency mocks, thread scheduling determines
which pipeline.run call lands first, so the competitor's call could
take index [0] and flake CI. Replace [0] indexing with a predicate
match on the canonicalized github_repos kwargs.
2. skills/last30days/scripts/last30days.py:662 - When auto_resolve returns
github_repos, it has already run canonicalize_github_repos(cap=5) and
ranked by relevance. The downstream block then re-canonicalized with
cap=None, which can re-sort by topic-slug match and clobber the
auto_resolve relevance order. Guard the second canonicalization with
a repos_from_auto_resolve flag so it only fires for user-supplied
--github-repo input.
Brave uses identical v10 AES-128-CBC encryption to Chrome; only the
DB path (BraveSoftware/Brave-Browser) and Keychain service name
("Brave Safe Storage") differ. Refactored chrome_cookies.py to share
a single _extract_chromium_cookies_macos helper rather than duplicating
the decryption logic.
Profile discovery tries Default/ first, then scans numbered Profile N/
directories so non-default Brave profiles are covered.
When yt-dlp is installed but stale (or otherwise unable to fetch transcripts
for any returned videos), runs previously reported YouTube as fully
successful in two user-facing surfaces:
1. Footer (render.py): showed "N videos | M views" with no indication
that zero transcripts were captured. The "with transcripts" segment
was conditionally suppressed when the count was zero - converting
the canonical stale-binary failure mode into a silent absence at
the very surface users read for "did this work?".
2. Quality nudge (quality_nudge.py): classified YouTube as "active"
based purely on yt-dlp installation + absence of a top-level error.
Per-video transcript-fetch ratio was never inspected. A run that
returned N videos with 0 transcripts (canonical stale-binary
failure) was reported as fully active.
The engine itself logs the failure correctly at default stderr level
(`[YouTube] Got transcripts for 0/N videos (N failed)`), but that line
gets buried in 100+ lines of parallel-source progress output and is
contradicted by the success-shaped footer and nudge that follow.
This change makes both conclusion surfaces honest:
* render.py footer always renders "M/N with transcripts" so the ratio
is visible regardless of value. Zero is no longer hidden. Format is
M/N (not bare M) so the denominator is in the message and the user
does not have to cross-reference the "videos" count.
* quality_nudge.py adds a third tier between "active" and "missing":
"degraded". Triggered when yt-dlp is installed AND videos were
returned AND transcript-fetch ratio is below threshold (default 50%,
tunable via DEGRADED_TRANSCRIPT_THRESHOLD env var). Emits an
actionable nudge: "YouTube returned N videos but only M transcripts
captured. The most common cause is a stale yt-dlp binary - YouTube's
caption format changes frequently and old binaries silently fail
every transcript. Update via your package manager: scoop update
yt-dlp (Windows), brew upgrade yt-dlp (macOS), or pip install -U
yt-dlp."
* last30days.py populates youtube_videos_count and
youtube_transcripts_count in the research_results dict it passes to
compute_quality_score, enabling the new degraded check at the call
site.
Threshold rationale: 50% accommodates a few legitimate
caption-disabled videos in a multi-video result, but a stale-binary
run that fails every transcript trips the nudge cleanly.
Score impact: degradation is informational, not score-affecting.
YouTube still counts as "active" in score_pct so users do not see
their score drop for a fixable client-side issue. The nudge directs
them to their own package manager.
Tests:
* tests/test_quality_nudge.py: 6 new TestYouTubeDegraded cases cover
zero-transcripts-flags-degraded, partial-above-threshold-does-not-flag,
zero-videos-does-not-flag (no false positives on absence),
one-of-three-flags-degraded, threshold-tunable-via-config, and
degraded-does-not-affect-score.
* tests/test_render_v3.py: 4 new YoutubeFooterTranscriptRatioTests
cases cover zero-transcripts-with-videos-renders-zero-over-total
(the regression repro), partial-renders-ratio, full-renders-ratio,
and no-videos-suppresses-entire-segment.
All 29 new test cases verified GREEN with the fix and RED without it
(temp-reverted both files separately to confirm each test catches the
specific regression it asserts).
Integration validation: ran the engine against an intentionally stale
yt-dlp 2025.03.31 binary placed first on PATH. Pre-fix the footer
showed `YouTube: 3 videos | 386,815 views` (no transcript signal).
Post-fix the footer shows `YouTube: 3 videos | 386,815 views | 0/3
with transcripts` and stderr emits "Degraded: YouTube" plus the
actionable update-yt-dlp nudge.
Out of scope (deserves its own PR): exposing transcripts_captured in
the EVIDENCE FOR SYNTHESIS block so the synthesizing model can flag
degradation in prose. Larger schema-touching change.
Asserts compute_save_path_display() never returns a backslash when the
save_dir is under the user's home directory, regardless of host OS.
Reproduces the original bug on Windows (failed message before the fix:
AssertionError: '\' unexpectedly found in
'~/l30d_save_path__luu2g76\Documents\Last30Days\british-airways-middle-east-raw-v3.md'
)
and locks in the contract on POSIX hosts (passes trivially today; would
fail if anyone removes .as_posix() in the future).
Verified by temporarily reverting the fix and confirming RED, then
re-applying the fix and confirming GREEN. All 13 CliV3Tests pass.
Trevin Chow