Install
$ agentstack add skill-eriemon-github-management-github-management ✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
About
Github Management
Use this skill for full-cycle GitHub repository management. Prefer gh CLI for authenticated operations and use GitHub REST or GraphQL only when gh does not expose the needed data cleanly.
Core Rule
Collect facts first, summarize risk, then ask before any mutating action.
Mutating actions include pushing commits, merging PRs, closing issues, deleting branches or tags, creating releases, changing labels or milestones, editing branch protection, changing repository settings, modifying permissions, dismissing reviews, rerunning workflows, and applying generated fixes.
Run the pipeline exactly:
Context -> Auth -> Read-only Inspect -> Risk Summary -> Confirmation -> Mutation -> Verification
If context, authentication, target identifiers, or user intent are missing, stop and ask. Do not guess.
Authentication Onboarding
When this skill is triggered, first decide whether GitHub authentication is already ready for the requested repo. If readiness is unknown or missing, tell the user how to configure GitHub access step by step before doing GitHub management work.
Do not skip this onboarding. Use references/authentication.md and present the user-facing steps there. Keep the guidance practical:
- Ask the user to create a classic Personal Access Token with
reposcope. - Tell them to save it locally under this skill's
config/folder, never in chat. - Tell them to run
gh auth login --with-token /scripts/inspect_repo_context.py" --repo "." --jsonwhen you need one read-only snapshot of local remotes, current branch, repo slug, and auth readiness.
- Identify the local repository path and remote with
git remote -v. - Confirm the target owner/repo with
gh repo view --json nameWithOwner,url. - If the task references a PR, issue, release, branch, label, or workflow, resolve its exact identifier before acting.
- Verify authentication.
- Read
references/authentication.mdand inspectconfig/auth.example.jsonfor the expected local auth shape. - Run
gh auth status. - If authentication or scope is missing, guide the user through `gh auth login --with-token /scripts/inspect_pr.py" --repo "." --json
python "/scripts/inspectci.py" --repo "." --json python "/scripts/inspectprchecks.py" --repo "." --json python "/scripts/inspectrepocontext.py" --repo "." --json python "/scripts/inspectrelease.py" --repo "." --json python "/scripts/fetchcomments.py" --repo "." --json python "/scripts/triageissues.py" --repo "." --json python "/scripts/repo_audit.py" --repo "." --json
All JSON output uses:
```json
{"ok": true, "source": "github-management", "items": [], "warnings": [], "next_actions": []}
Task Guide
| Task | First inspection | Confirmation needed before | | --- | --- | --- | | Issue triage | triage_issues.py or gh issue list --json ... | labeling, assigning, closing, editing milestone | | Repo context | inspect_repo_context.py | switching target repo assumptions, pushing, merging, release edits | | PR review | inspect_pr.py | pushing fixes, submitting reviews, merging | | Review comments | fetch_comments.py; read references/review-comments.md | resolving conversations or pushing fixes | | CI failure | inspect_ci.py or inspect_pr_checks.py; read references/ci-diagnostics.md | rerunning workflows or applying fixes | | Release | inspect_release.py or gh release list/view | creating, editing, deleting, or publishing a release | | Repo audit | repo_audit.py | changing settings, protections, permissions, secrets, Actions policy | | Security review | Read the matching security best-practice reference listed below | filing issues, changing policy, applying fixes | | Threat model | Read references/security-threat-model-template.md and references/security-controls-and-assets.md | writing files, filing issues, applying mitigations | | Security ownership map | Run scripts/security_ownership/run_ownership_map.py and query with scripts/security_ownership/query_ownership.py | writing analysis outside the requested output directory |
Security Reference Routing
For security best practices, identify the repo language and framework, then load every matching reference:
- Go backend:
references/security-best-practices-golang-general-backend-security.md - JavaScript Express backend:
references/security-best-practices-javascript-express-web-server-security.md - JavaScript frontend, no framework or unclear framework:
references/security-best-practices-javascript-general-web-frontend-security.md - jQuery frontend:
references/security-best-practices-javascript-jquery-web-frontend-security.md - TypeScript/Next.js backend or full-stack app:
references/security-best-practices-javascript-typescript-nextjs-web-server-security.md - TypeScript/React frontend:
references/security-best-practices-javascript-typescript-react-web-frontend-security.md - TypeScript/Vue frontend:
references/security-best-practices-javascript-typescript-vue-web-frontend-security.md - Python Django backend:
references/security-best-practices-python-django-web-server-security.md - Python FastAPI backend:
references/security-best-practices-python-fastapi-web-server-security.md - Python Flask backend:
references/security-best-practices-python-flask-web-server-security.md
For threat modeling, stay repo-grounded: use evidence anchors, trust boundaries, assets, attacker capabilities, abuse paths, mitigations, and explicit assumptions from references/security-threat-model-template.md.
For security ownership map tasks, keep analysis read-only against git history. Write artifacts only to the user-specified output directory or a repo-local ownership-map-out/, then use references/security-ownership-neo4j-import.md only when graph import is requested.
References
- Read
references/authentication.mdbefore configuring or troubleshooting GitHub authentication. - Read
references/safety-policy.mdbefore any mutating GitHub action. - Read
references/workflows.mdfor task-specific GitHub workflows. - Read
references/review-comments.mdfor PR comment and review-thread handling. - Read
references/ci-diagnostics.mdfor failing check and log inspection. - Read
references/development-practices.mdbefore code changes caused by GitHub management work. - Read
references/release-installation.mdfor this skill's local release branch, floating install tag, and Codex install path. - Read
references/source-attribution.mdwhen source provenance matters.
Common Mistakes
- Do not infer authentication from local git access; always check
gh auth status. - Do not treat external CI as GitHub Actions; report the URL and ask for access or logs.
- Do not merge or close anything because tests pass; summarize and ask first.
- Do not modify repository settings from an audit command. Audits are read-only.
- Do not assume
releaseis a safe short ref if a tag namedreleaseexists; usegithub-management-releasefor the floating install tag and report ambiguity explicitly. - Do not hide uncertainty. If GitHub returns partial data or permissions block a check, report the gap.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: Eriemon
- Source: Eriemon/github-management
- License: Apache-2.0
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet — be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.