You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
📰 EU Parliament Committee Reports — Unified Workflow
🔖 Workflow Parameters
Parameter
Value
ARTICLE_TYPE_SLUG
committee-reports
Family
Unified (Stages A → B → C → D → E in one workflow)
Data window
last 7 days
Primary feeds
get_committee_documents, get_committee_documents_feed, get_procedures_feed, get_events_feed with timeframe: "one-week".
Stage A budget
≤ 4–5 min (per article-horizons.ts)
Stage B budget (2 passes)
22–28 min — HARD CEILING per article-horizons.ts (do not exceed the per-slug ceiling on Stage B even if Pass 2 still has shallow sections; force GATE_RESULT=ANALYSIS_ONLY instead)
Stage C budget (gate + optional Pass 3)
≤ 4 min; stop repair diagnostics at minute 32 and ship ANALYSIS_ONLY if still not GREEN
Stage D budget
≤ 2 min (deterministic)
Stage E budget (commit + single PR)
≤ 2 min
Stage C exit tripwire
minute 36 elapsed (long-horizon prospective: 39; long-horizon retrospective: 38; electoral: 42) — the decision threshold for forcing GATE_RESULT=ANALYSIS_ONLY and (if late) skipping Stage D so the run can still reach the PR call. Per-slug stage ceilings live in src/config/article-horizons.ts; the tripwire backstops any per-stage overrun. Note: Stage D + E run after this tripwire, between the Stage C exit and the PR-call deadline.
Hard PR-call deadline
minute ≤ 45 elapsed (target ≤ 42) — deadline for the single safe-outputs create_pull_request call. Note: engine.mcp.session-timeout is intentionally NOT set — gh-aw v0.71.3 advertises this field but the bundled gateway image (v0.3.1) rejects it; the agent must finish within the 60-min timeout-minutes cap regardless.
Ground every claim in a named committee + document ID (e.g. ENVI draft report on directive 2025/0042(COD)).
Include existing/committee-productivity.md in the analysis set.
🧯 Committee-reports invocation-cap guard (run #25715099069):
Copilot currently enforces a 100 model invocation cap per agent session; this
workflow previously hit that cap while repeatedly inspecting validator output
and source code during Stage C. For committee-reports, Stage C therefore has
a local repair cutoff at minute 32 elapsed (four minutes before the
standard minute-36 Stage C exit tripwire). At the start of each Stage C
boundary — before the first validator run, after the first RED result, and
after any Pass 3 edits — recompute ELAPSED_MIN. If ELAPSED_MIN ≥ 32, set
GATE_RESULT=ANALYSIS_ONLY, emit the ANALYSIS_ONLY gate line, skip all
remaining validator/debug/source-inspection commands, and proceed directly to
Stage E. Stage C may run the validator at most twice total: one initial run,
one post-Pass-3 rerun. Never use extra grep, sed, node, or source-code
probes to interpret validator failures inside Stage C; use only the validator
output already produced. A timely analysis-only PR is preferred over exhausting
the invocation cap before the PR call.
Stage A — Data Collection (Ref: 01, 07)
Run the canonical gateway block from 08-infrastructure.md §4. Source
scripts/mcp-setup.sh, then scripts/wb-mcp-probe.sh and
scripts/imf-mcp-probe.sh. Collect EP feed data first; fall back to
direct endpoints on failure. Deep-fetch up to 10 procedures / voting
records / meeting decisions into ${ANALYSIS_DIR}/data/. Target ≤ 4 min.