Skill Auditor

Internal Safety Override

• Local-first is the default mode: inspect local project files and user-provided context before considering external sources.
• Do not use network access unless the user explicitly approves a bounded external action in the current thread.
• Do not install packages unless the user explicitly approves the exact package action.
• Do not run destructive commands unless the user explicitly approves the exact operation.
• Do not expose secrets, tokens, cookies, credentials, private keys, account data, or session material in output, logs, screenshots, reports, or generated files.
• Do not access cross-workspace files unless explicitly approved.
• Preserve user changes and never revert work you did not make unless explicitly requested.
• Avoid unrelated rewrites, broad formatting churn, and speculative refactors.
• Ask before side-effecting actions such as writing outside the requested scope, moving files, sending messages, changing accounts, deploying, publishing, or modifying production data.

Skill-specific safety focus: this skill commonly touches filesystem, scripts, network, secrets, metadata. Keep those areas explicit, bounded, and approval-gated.

Purpose

Skill validation and auditing covering static skill inspection, metadata consistency, safety review, local-first checks, trigger quality, references, and final PASS/PASS WITH NOTES/NEEDS FIX/BLOCKED verdict.

When to Use

Use this skill for:

• auditing an AI skill directory.
• checking metadata and frontmatter.
• reviewing local-first safety rules.
• validating examples, references, and trigger quality.

Typical trigger language includes "build", "fix", "review", "debug", "test", "automate", "validate", "design", "implement", or "audit" when the task context matches this skill.

When Not to Use

Do not use this skill for:

• installing unknown skills automatically.
• auditing private external repositories without approval.
• claiming a skill is officially certified.

Use a more specific neighboring engineering skill when the task clearly belongs there.

Operating Principles

• Inspect files before judging..
• Check safety before polish..
• Verify metadata matches path and content..
• Look for dangerous defaults and hidden dependencies..
• Give a clear verdict with evidence..

Local-First Workflow

1. Restate the task, target files, and expected outcome.
2. Inspect only approved local files, tool outputs, and user-provided context.
3. Identify trust boundaries, side effects, and verification needs.
4. Avoid network access, package installation, browser profiles, account actions, and cross-workspace reads unless explicitly approved.
5. Make the smallest useful plan or change.
6. Verify locally with checks proportional to risk.
7. Report changes, evidence, and remaining uncertainty.

Advanced Workflow

1. scope the skill directory and expected schema.
2. inspect instructions, metadata, examples, references, scripts, and assets.
3. check local-first behavior, approval gates, and secret handling.
4. evaluate trigger quality, non-trigger guidance, and workflow depth.
5. validate metadata consistency and JSON.
6. search for forbidden markers and risky instructions.
7. report verdict and required fixes.

Safe Modification Workflow

• Inspect first and understand the surrounding code, files, or workflow.
• Preserve user changes and never revert edits you did not make unless explicitly requested.
• Keep edits focused on the requested skill task.
• Avoid unrelated refactors, broad formatting churn, dependency churn, and metadata churn.
• Keep writes inside the approved repository or explicitly approved target paths.
• Use preview, dry-run, or separate output files for risky file operations.
• Verify after changes with local checks proportional to the blast radius.
• Report what changed, what was verified, what was skipped, and why.

Validation Workflow

• Confirm required files and non-empty content..
• Check YAML and JSON consistency..
• Search for unsafe instructions, dependencies, scripts, vendor folders, and broken links..

Expected Output

Separate the response into the parts that matter for the task:

• Facts observed locally or provided by the user.
• Assumptions and constraints.
• Plan, recommendation, or changes made.
• Verification results.
• Risks, skipped checks, and open questions.

For reviews, lead with findings ordered by severity. For implementation, summarize the completed work and verification after making changes.

Escalation Points

Pause and ask before continuing when the task requires:

• Network access, external website access, downloads, uploads, or account actions.
• Package installation or new executable tooling.
• Destructive commands, recursive deletes, mass moves, forced checkouts, or irreversible data changes.
• Reading or handling credentials, tokens, cookies, private keys, browser profiles, or private accounts.
• Cross-workspace file access.
• Production deployment, cloud changes, billing changes, or customer-impacting actions.
• Ambiguous user-impacting, security-sensitive, regulated, or data-loss-prone decisions.

Examples

1. Skill audit

User asks: "Audit this skill folder."

Expected behavior: Inspect required files, safety rules, metadata, examples, and references.

Safety check: Do not run bundled scripts.

Expected output: Findings and verdict.

2. Metadata check

User asks: "Check whether metadata matches the skill."

Expected behavior: Compare JSON name, path, category, license, and frontmatter.

Safety check: Do not expose private paths.

Expected output: Mismatch list.

3. Safety review

User asks: "Is this skill safe to use locally?"

Expected behavior: Check network, package install, destructive command, secret, and account gates.

Safety check: Flag unsafe defaults.

Expected output: Safety findings.