Pattern miner: Add cross-repository monitor archetype from contribution-check pattern

[github-actions]

Caution

Protected files were modified in this change.
This pull request is in request_review mode and requires explicit human scrutiny before merge.

Protected files: .github/copilot-instructions.md

Mined pattern: cross-repository monitor/orchestrator

Upstream evidence

• githubnext-agentics/files/workflows/contribution-check.md — a schedule-triggered orchestrator that reads PRs from ${{ vars.TARGET_REPOSITORY }}, dispatches each PR to a stateless sub-agent (.github/agents/contribution-checker.agent.md) for evaluation, compiles results into one report issue in the home repo, and writes add-labels/add-comment back to the target repo using target-repo:.
• github-gh-aw/files/.github/aw/workflow-patterns.md (lines 284-293) — the "Cross-Repository Pattern" section explicitly documents this as a recommended architecture: enable the needed GitHub toolsets for the external repo, configure PAT/GitHub App auth for cross-repo writes in safe-outputs:, set target-repo explicitly, and document required token scopes in the prompt.

One upstream workflow demonstrates the shape, and upstream documentation explicitly describes it as a recommended pattern, satisfying the mining threshold.

Why the library did not already cover it

• No existing archetype's recommended_triggers/recommended_safe_outputs represented reading one repository and writing to another with a target-repo-based split (checked via a text search of current-library.json for target-repo, TARGET_REPOSITORY, cross-repo, orchestrator).
• contribution-guidelines-checker has a one-line tip mentioning target-repo as an aside, but no archetype is built around the monitor/orchestrator shape itself (pre-filter → dispatch → compile report → write back cross-repo).

What changed in patterns/

• Added patterns/archetypes/cross-repo-monitor.json (new curated archetype, success_rate: null, count: 0) describing the schedule-triggered monitor/orchestrator shape, with safe outputs issues, add-labels, add-comment, and tips covering target-repo, pre-filtering in steps:, per-item dispatch, report compilation with close-older-issues, and token-scope documentation.
• Appended cross-repo-monitor to archetypes in patterns/manifest.json.
• Added the matching runtime entry to patterns/workflow-generation.json (icon, capabilities, instructions: [workflow-patterns.md], body).
• Updated .github/copilot-instructions.md: bumped the archetype count to 34 and added cross-repo-monitor to the curated-archetypes list, so test/copilot-instructions.test.js stays in sync.

Small refinement

• patterns/archetypes/issue-hierarchy-manager.json: added one tip about create-issue's group: true option (groups several new findings as linked sub-issues under one parent). Evidenced by two upstream workflows — githubnext-agentics/files/workflows/issue-arborist.md (group: true with max: 5) and githubnext-agentics/files/workflows/duplicate-code-detector.md (group: true with max: 3) — plus the documented option in github-gh-aw/files/.github/aw/safe-outputs-content.md (line 102). No existing archetype's tips mentioned group: true.

Validation

• npm test: 19 test files, 329 tests passed (including test/copilot-instructions.test.js and test/patterns.test.js, which assert every manifest archetype id has a matching file and is listed in the instructions).
• npm run build: succeeded; dist/patterns/archetypes/cross-repo-monitor.json is emitted alongside the other archetype files.

Candidates deliberately left out

• A rotating "focus area" dimension (seen in repository-quality-improver.md, efficiency-improver.md, tech-content-editorial-board.md) is already captured via the cache-memory rotation tips on code-health-auditor and performance-nut.
• The open-PR-cap pre_activation gate (11 upstream workflows) is already present as a tip on code-improvement, documentation-updater, daily-test-improver, repo-maintainer, performance-nut, and link-checker.
• A dedicated "discussion task miner" archetype (discussion-task-miner.md) was considered but only one upstream workflow demonstrates it with no corroborating documented pattern beyond the general issue/discussion safe-output reference, so it did not meet the two-workflow-or-documented-pattern bar as a standalone archetype.
• A "wiki writer/coder" pair (agentic-wiki-writer.md, agentic-wiki-coder.md) uses the gollum trigger and repo-memory, but gollum is not documented as a supported trigger anywhere in the mined github-gh-aw docs, so it was left out pending confirmed trigger support.

Generated by Pattern Miner · copilot · auto · 311.5 AIC · ⌖ 17.7 AIC · ⊞ 8.3K · ◷

[github-actions]

Protected files were modified in this pull request and require manual scrutiny before merge.

Please verify that each protected-file change is intentional, policy-compliant, and safe:

• Protected files: .github/copilot-instructions.md

[github-actions]

Reviewer dispatch: auto → both (docs change in .github/copilot-instructions.md + structured pattern data in patterns/*.json, no TS/JS/CSS code touched).

Reviewed the new cross-repo-monitor archetype, the issue-hierarchy-manager tip addition, and the manifest/instructions updates.

Findings: no blocking issues.

• All modified/added JSON (manifest.json, workflow-generation.json, cross-repo-monitor.json, issue-hierarchy-manager.json) parses correctly and follows the existing archetype schema (id, recommended_triggers, recommended_safe_outputs/recommended_tools, permissions, capabilities, body template placeholders).
• .github/copilot-instructions.md count bump (33→34) and curated-archetype list addition match the new manifest entry, consistent with what test/copilot-instructions.test.js and test/patterns.test.js assert (archetype count, curated vs. empirical listing, every manifest id having a matching archetype file).
• The new archetype's body template references target-repo safe outputs in prose ("pre-configured target-repo safe outputs"), but — same as the existing contribution-guidelines-checker precedent — the generator (src/js/workflow.js) has no logic that actually injects target-repo: into generated safe-output blocks; it's left as a manual follow-up, matching the pattern already established elsewhere in this codebase rather than a regression introduced here.
• icon: "repo" and prompt_style: "role-steps" reuse existing conventions; no new vocabulary introduced.

Nothing here warrants REQUEST_CHANGES. (Note: this PR shows as already merged at review time.)

Warning

Firewall blocked 2 domains

The following domains were blocked by the firewall during workflow execution:

• github.com
• registry.npmjs.org

To allow these domains, add them to the network.allowed list in your workflow frontmatter:

network:
  allowed:
    - defaults
    - "github.com"
    - "registry.npmjs.org"

See Network Configuration for more information.

Generated by Specialist PR Review for #320 · copilot · auto · 95.5 AIC · ⌖ 7.4 AIC · ⊞ 7.9K