Add github-repository-inventory skill 🤖🤖🤖

[Kinosaur]

A read-only, privacy-first Agent Skill that answers one question the platform does not make easy and models do unreliably: what has this GitHub account actually contributed to — across repos it does not own — and what is the evidence? It models the user's relationship to each repo (owner / collaborator / org-member / PR author / PR reviewer / issue author / commit-contributor) with the evidence and a confidence level, and never conflates repository access with verified contribution.

Fills a gap in the collection

Existing skills operate on a single checked-out repo (e.g. acquire-codebase-knowledge, repo-story-time); none work at the account level across owned and external repositories with a relationship-evidence model. This is a distinct capability, not a duplicate.

Why this is meaningful uplift (not something frontier models already do well)

• Relationship evidence + confidence. Every relationship is recorded with the evidence that established it and a confidence level. Access is never reported as contribution — a distinction most tools conflate.
• Privacy-first by construction. Public-only and read-only by default; no token storage; private repos require an explicit flag, have README content redacted, and are written only to a gitignored location (with a guard that warns otherwise). Private repo names never reach shareable output.
• Deterministic, testable core. Identical input produces byte-identical output. All GitHub access goes through gh; the transform/render core is pure and unit-tested against recorded fixtures (no live API in tests).
• Honest about gaps. Deleted repos, expired permissions, the commit-contribution time window, and rate-limit truncation are surfaced in warnings.json rather than hidden.

What it produces

catalog.json (canonical), PROJECTS.md (human-readable, localizable — English/Burmese), warnings.json, and optional derived reports (technology summary, portfolio candidates, missing READMEs, contribution summary).

Best for

Reconstructing a verifiable contribution history (job applications, reviews, proving cross-org/contract work) and giving an agent a truthful tool instead of letting it guess. A focused, occasional-use utility — and honest about that.

Requirements

Python 3.11+ and an authenticated GitHub CLI (gh) 2.90.0+. Bundles a small standard-library Python program under scripts/ (no third-party packages; gh owns auth). Read-only; never modifies repositories.

Upstream source & full test suite (80+ tests): https://github.com/Kinosaur/github-repository-inventory

---

🤖 Authored with Claude Code.

[github-actions]

🔍 Skill Validator Results

✅ All checks passed

Scope	Checked
Skills	1
Agents	0
Total	1

Severity	Count
❌ Errors	0
⚠️ Warnings	0
ℹ️ Advisories	0

Summary

Level	Finding
ℹ️	Found 1 skill(s)
ℹ️	[github-repository-inventory] 📊 github-repository-inventory: 1,204 BPE tokens [chars/4: 1,379] (detailed ✓), 15 sections, 1 code blocks
ℹ️	✅ All checks passed (1 skill(s))

Full validator output

Found 1 skill(s)
[github-repository-inventory] 📊 github-repository-inventory: 1,204 BPE tokens [chars/4: 1,379] (detailed ✓), 15 sections, 1 code blocks
✅ All checks passed (1 skill(s))

[Copilot]

Pull request overview

Note

Copilot was unable to run its full agentic suite in this review.

Introduces a new github-repository-inventory skill that scans an authenticated GitHub account (via gh) to produce a deterministic, privacy-first inventory of owned, contributed-to, and optionally accessible/private repositories, plus derived reports and schema validation.

Changes:

• Add end-to-end inventory CLI (scan/discover/inspect/render/report/validate) that discovers repos via REST/GraphQL/Search, inspects README + top-level contents, and writes catalog.json, PROJECTS.md, and warnings.json.
• Add pure technology detection and project-type classification based on root manifests + dependency parsing.
• Add documentation/reference materials and register the skill in docs/README.skills.md.

Reviewed changes

Copilot reviewed 9 out of 9 changed files in this pull request and generated 5 comments.

Show a summary per file

File	Description
skills/github-repository-inventory/scripts/inventory.py	Implements the CLI workflow, discovery/inspection pipeline, deterministic rendering, caching, and schema validation.
skills/github-repository-inventory/scripts/detect_technologies.py	Adds pure manifest detection + dependency parsing to emit evidence-backed technologies and classify project types.
skills/github-repository-inventory/references/repository-schema.md	Documents the repository record/catalog schema for humans.
skills/github-repository-inventory/references/relationship-model.md	Documents relationship types/evidence/confidence model used in outputs.
skills/github-repository-inventory/references/privacy-rules.md	Documents privacy guarantees and safe handling rules for private data.
skills/github-repository-inventory/references/limitations.md	Documents known coverage/inspection limitations surfaced in warnings.
skills/github-repository-inventory/assets/projects-template.md	Provides a PROJECTS.md template reference for the skill.
skills/github-repository-inventory/SKILL.md	Defines the skill metadata, safety defaults, and usage workflow/commands.
docs/README.skills.md	Registers the new skill in the skills index.