om-auto-sec-report

Aggregates per-unit security analyses across merged PRs into one docs-only report PR.

1.7k|382|Updated Sep 10, 2025
One-click install
npx skills add https://github.com/open-mercato/open-mercato --skill om-auto-sec-report
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: om-auto-sec-report
Source: https://github.com/open-mercato/open-mercato/tree/main/.ai/skills/om-auto-sec-report
Command: npx skills add https://github.com/open-mercato/open-mercato --skill om-auto-sec-report

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Reviewing security posture across many merged pull requests one at a time is slow and easy to skip. This Skill automates the batch analysis by looping a single-unit security review skill over a window of work (a date range, PR-number floor, branch, or spec) and consolidating every finding into one aggregated markdown and HTML report delivered as a docs-only PR against develop.

Core Features & Use Cases

  • Window-based batching: Analyze all PRs merged since a date, above a PR number, or default to the last 7 days of merged PRs, with --max-units caps and --include-open support.
  • Delegated per-unit analysis: Invokes om-auto-sec-report-pr for each unit with --out-fragment, then concatenates fragments verbatim into .ai/analysis/auto-sec-report-{DATE}.md plus an HTML mirror.
  • Consolidated reporting: Produces an executive summary, OWASP risk heatmap, paranoid-vector coverage matrix, and a deduplicated "Next steps — go deeper" list with exactly one [recommended] follow-up.
  • Resumable orchestration: Tracks progress in a plan file with per-unit checkboxes and commit SHAs, hands off to om-auto-continue-pr when a run cannot finish in one invocation.
  • Use Case: After a sprint, run the driver with no arguments to audit the last 7 days of merged PRs and receive a single PR containing the full security report, labeled review, documentation, security, and skip-qa.

Quick Start

Run the auto security report driver with no arguments to analyze the last 7 days of merged PRs and open an aggregated docs-only report PR against develop.

Frequently Asked Questions about om-auto-sec-report

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run a security review across multiple merged pull requests?

Invoke the driver with a date (YYYY-MM-DD), a PR-number floor, or no argument to default to the last 7 days of merged PRs. It queues each unit, runs the per-unit analysis skill on each, and opens one aggregated docs-only PR against develop.

What does the aggregated security report contain?

The report includes an executive summary with finding counts by severity, an OWASP A01-A10 risk heatmap, a paranoid-vector coverage matrix, verbatim per-unit fragments, and a deduplicated Next steps list with one recommended follow-up. It is written as markdown and HTML under .ai/analysis/.

Can I analyze a single branch or spec instead of merged PRs?

Yes. Passing a branch name or a spec path under .ai/specs/ creates a single-entry queue, and the driver still produces the full aggregate report layout. Branch and spec inputs are first-class window types.

What happens if the batch security run cannot finish in one invocation?

The driver leaves Status: in-progress in the PR body, keeps per-unit progress checkboxes with commit SHAs, and posts a hand-off comment directing you to resume with /auto-continue-pr. Partial or skipped units are recorded inline without aborting the batch.

Does the security report ever include secrets or sensitive data?

No. The skill forbids pasting raw diffs, secrets, tokens, .env content, credentials, internal hostnames, or user PII, redacting them to {REDACTED}. A secret-leak grep and link validation run before any commit.