Description
The Daily Regulatory Report (discussion #55519, 2026-08-24) found its github_discussion_query mcp-script tool fails with "Argument list too long" for --limit values above ~11-12, which the report attributes to jq being invoked with large heredoc/<<< args. This capped that run's review to the 10 most-recently-updated discussions instead of a full 24-48h scan by creation date, meaning several expected daily reports (Issues Report, Firewall Report, Token Consumption Report, Team Status) were silently missed from cross-checking simply because they fell outside the truncated sample. The likely location is the discussion-pagination path in .github/workflows/shared/github-queries-mcp-script.md (the github-discussion-query tool), which passes accumulated JSON through jq ... <<< "$OUTPUT" here-strings as data grows.
Expected Impact
Fixes a tooling limitation that's silently shrinking the Daily Regulatory Report's effective review coverage — right now it can only reliably sample ~10-12 discussions per run regardless of how many actually need cross-checking.
Suggested Agent
New Agent — investigate the jq/heredoc argument-passing in the discussion-query path of .github/workflows/shared/github-queries-mcp-script.md and switch to streaming input (e.g. jq reading from a temp file or stdin pipe) instead of large here-string/heredoc arguments, so --limit (or since-based pagination) scales past ~12 items.
Estimated Effort
Quick (< 1 hour) — likely a small refactor from <<< here-strings to a piped/temp-file jq invocation.
Data Source
DeepReport Intelligence Briefing analysis, 2026-08-25, sourced from Daily Regulatory Report discussion #55519 (Data Quality Notes / Workflow Suggestions).
Generated by 🔬 Deep Report · claude · agent · 249.5 AIC · ⌖ 8.79 AIC · ⊞ 12.4K · ◷
Description
The Daily Regulatory Report (discussion #55519, 2026-08-24) found its
github_discussion_querymcp-script tool fails with "Argument list too long" for--limitvalues above ~11-12, which the report attributes tojqbeing invoked with large heredoc/<<<args. This capped that run's review to the 10 most-recently-updated discussions instead of a full 24-48h scan by creation date, meaning several expected daily reports (Issues Report, Firewall Report, Token Consumption Report, Team Status) were silently missed from cross-checking simply because they fell outside the truncated sample. The likely location is the discussion-pagination path in.github/workflows/shared/github-queries-mcp-script.md(thegithub-discussion-querytool), which passes accumulated JSON throughjq ... <<< "$OUTPUT"here-strings as data grows.Expected Impact
Fixes a tooling limitation that's silently shrinking the Daily Regulatory Report's effective review coverage — right now it can only reliably sample ~10-12 discussions per run regardless of how many actually need cross-checking.
Suggested Agent
New Agent — investigate the
jq/heredoc argument-passing in the discussion-query path of.github/workflows/shared/github-queries-mcp-script.mdand switch to streaming input (e.g.jqreading from a temp file or stdin pipe) instead of large here-string/heredoc arguments, so--limit(orsince-based pagination) scales past ~12 items.Estimated Effort
Quick (< 1 hour) — likely a small refactor from
<<<here-strings to a piped/temp-filejqinvocation.Data Source
DeepReport Intelligence Briefing analysis, 2026-08-25, sourced from Daily Regulatory Report discussion #55519 (Data Quality Notes / Workflow Suggestions).