- Home
- Skills
- DevOps & Cloud
- public mirror secret guard
Works with the AI Agents you already use
public mirror secret guard
Automated secret scanning and identity scrubbing for repositories mirrored to public GitHub.
Free
public mirror secret guard
Example session with this skill installed
Audit my staged changes for a public mirror push. I have an updated .env and some Kubernetes values.yaml changes.
- Read your context and instructions
- Compiled the public mirror secret
I have audited your staged changes using the scan_diff_guard.py script. The scan identified one missing configuration counterpart and verified that no internal hostnames were leaked in the diff.
Security Audit Report
| File Path | Status | Action Taken |
|---|---|---|
infrastructure/helm/values.yaml | PASS | No internal hostnames or secrets detected in diff. |
.env | IGNORED | File is correctly listed in .gitignore. |
.env.example | WARNING | Missing key STRIPE_SECRET_KEY found in .env. |
PROMPTLOG.md | PASS | No verbatim credentials found in markdown logs. |
AGENTS.local.md | PROTECTED | Machine state correctly isolated from tracked AGENTS.md. |
Detected Counterpart Mismatch
The following keys exist in your git-ignored .env but are missing from .env.example:
STRIPE_SECRET_KEYDB_INTERNAL_URL
Next steps
- Sync
.env.exampleby adding the missing keys with placeholder values. - Move the cluster endpoint notes from
README.mdtoAGENTS.local.mdto ensure they stay off the mirror. - Run
uv run scripts/scan_diff_guard.py --stagedagain to confirm a clean bill of health before pushing.
Connects securely to your tools. The creator never sees your data.
About this skill
Managing a public repository mirror is high risk. One accidental commit of a Kubernetes secret or internal hostname creates a permanent security hole in your Git history. This skill enforces pre-commit hygiene to ensure sensitive data stays local and public mirrors remain clean.
What it does
- Diff-level scanning audits
git diff --cachedto catch secrets before they are committed, not just in the final file state. - Identity scrubbing removes upstream boilerplate slugs and brand names to prevent leaking project origins.
- State separation mandates the use of git-ignored
*.local.mdfiles for machine-specific notes and cluster paths. - Counterpart parity ensures every ignored
.envorvalues.yamlfile has a matching.exampletemplate with identical keys. - Fail-closed enforcement halts the workflow immediately if suspicious tokens or internal hostnames are detected.
How it works
- Run the
scan_diff_guard.pyscript against staged changes to identify leaked credentials or banned strings. - Generate
.examplefiles for all git-ignored configurations to maintain schema parity without exposing values. - Migrate environment-specific deployment notes from tracked files into
@AGENTS.local.md. - Configure the
--banned-brandflag to filter out upstream template identifiers during scaffolding.
Frameworks & tools
This skill integrates with Git for diff analysis, uses uv for script execution, and manages configuration files like values.yaml, .env, and Markdown logs.
Why this beats prompting it yourself
Manual prompts often miss secrets hidden in large diffs or forgotten upstream boilerplate references. This skill provides a systematic, script-driven audit that enforces file-level architectural rules like .local.md separation that LLMs usually overlook.
Use cases
- Auditing staged changes before pushing a private repo to a public GitHub mirror.
- Scrubbing brand names and internal emails when scaffolding a new project from a boilerplate template.
- Maintaining
.env.examplefiles automatically as local environment variables evolve. - Preventing
PROMPTLOG.mdor Kubernetes manifests from leaking internal infrastructure hostnames.
Known limitations
Requires uv to be installed for script execution. Scanning is limited to staged changes and does not retroactively clean existing Git history.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for DevOps & Cloud.
Requires the uv package manager. Only scans staged changes via git diff --cached. Does not perform retroactive history cleaning of already pushed commits.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free to use
No account required to browse
Trust & safety
Security scan
Security scan passed 4 days ago
- Free with an Agensi account
- Software
Python
Git