Install
$ agentstack add skill-aliengiraffe-vigilante-vigilante-issue-implementation-on-github-actions ✓ 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.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
Vigilante GitHub Actions Issue Implementation
Focus
- Read the prompt for detected tech stacks, process hints, and security guidance before changing workflow files.
- Keep changes scoped to the issue and do not broaden into unrelated workflow or repository changes.
- Treat
.github/workflows/as a security-sensitive surface. Every workflow edit should consider permissions, secret exposure, and supply-chain risk.
Workflow File Conventions
- Use
.ymlor.yamlconsistently with the repository's existing convention. Do not mix extensions within the same repository. - Validate workflow syntax before committing. Use
actionlintwhen it is available in the repository or installed locally. Ifactionlintis not available, note its absence and continue — do not fabricate output. - Keep workflow files readable: use clear job and step names, add inline comments for non-obvious logic, and prefer reusable workflows or composite actions over duplicated step blocks.
Pinned Actions
- Pin third-party actions to full commit SHAs, not mutable tags or branch references. Example:
uses: actions/checkout@with a trailing version comment. - When updating an action version, verify the new SHA corresponds to a reviewed release or tag.
- First-party GitHub actions (
actions/*) should also be pinned to SHAs for consistency and supply-chain safety. - When adding a new third-party action, prefer well-maintained actions with high community adoption. Avoid actions that request broad permissions or lack clear provenance.
Least-Privilege Permissions
- Always declare a top-level
permissions:block in workflow files. Default to the most restrictive set needed. - Use read-only
contents: readunless the workflow must write (e.g., creating releases, pushing tags, commenting on PRs). - Scope token permissions per job when different jobs need different access levels.
- Never use
permissions: write-allor leave permissions unspecified, which defaults to broad access in some repository configurations.
Secret and Credential Safety
- Never echo, log, or interpolate secrets directly in
run:shell commands. Pass secrets through environment variables. - Use
::add-mask::to mask dynamic values that may appear in logs. - Prefer OIDC-based authentication (e.g.,
aws-actions/configure-aws-credentialswithrole-to-assume) over long-lived cloud credentials stored as repository secrets. - Do not store secrets, tokens, or credentials in workflow files or committed configuration.
- When a workflow needs elevated access, document why in a comment and scope the access as narrowly as possible.
Safe Workflow Authoring
- Never interpolate untrusted event data (such as
${{ github.event.pull_request.title }}or${{ github.event.issue.body }}) directly intorun:shell scripts. Use an intermediate environment variable to prevent script injection. - Prefer
pull_requestoverpull_request_targetunless cross-fork access is explicitly required and the workflow is hardened against injection. - Use
concurrencygroups to prevent redundant or conflicting workflow runs. - Set appropriate
timeout-minuteson jobs to prevent hung runners from consuming resources.
Reusable Workflows and Composite Actions
- Prefer the repository's existing reusable workflows and composite actions over duplicating logic.
- When creating new reusable workflows, define clear
inputsandsecretscontracts. - Respect the repository's branch-protection rules and required status checks when adding or modifying workflows.
Mixed-Stack Repositories
- A repository with GitHub Actions workflows often also contains application code in Go, Node.js, Python, or other languages.
- Scope workflow-specific guidance to
.github/workflows/and related CI/CD configuration only. Do not apply workflow linting or hardening rules to application source code. - When an issue touches both workflow files and application code, validate each side with its appropriate toolchain.
- Check the prompt for additional detected tech stacks and follow their respective guidance for non-workflow changes.
Workflow
- Follow the base
vigilante-issue-implementationworkflow for issue comments, validation, push, and PR creation, including stacked base-branch detection (Base branch:directive in the issue body). - Use
vigilante commitfor all commit-producing operations. Do not usegit commitor GitHub CLI commit flows directly. - Any commit or amend must preserve the user's existing git author, committer, and signing configuration. Commit on behalf of the user and do not overwrite
git configwith a coding-agent identity. - Do not add
Co-authored by:trailers or any other agent attribution for Codex, Claude, Gemini, or similar coding-agent identities. - Repository-specific instructions (
AGENTS.md,README.md, CI config) remain authoritative when they are more specific than the generic GitHub Actions guidance in this skill.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: aliengiraffe
- Source: aliengiraffe/vigilante
- 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.