Skip to content
Back to skills

Ashrafiucse Secrets Detection

ASecurity

Finds hardcoded secrets in any project — API keys (AWS, GitHub, Google, Stripe, Slack), passwords, tokens, private keys, JWTs, database URLs with credentials. Runs a fast regex scan, then triages matches to remove false positives and checks .env/git hygiene. Use when auditing a codebase for leaked credentials or before committing/publishing a repo.

  • 76 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added September 22, 2026
securitygobashawsgitapidatabasesecurity

Works with

  • vscode
  • api

Security analysis

A100/100

Pro scans all 7 files and shows the line behind each finding

Scanned September 23, 2026

npx -y skills add jiayaoqijia/cryptoskill --skill ashrafiucse-secrets-detection --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Ashrafiucse Secrets Detection?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Ashrafiucse Secrets Detection
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/jiayaoqijia-ashrafiucse-secrets-detection/badge)](https://www.skillsdirectory.com/skills/jiayaoqijia-ashrafiucse-secrets-detection)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: secrets-detection
description: Finds hardcoded secrets in any project — API keys (AWS, GitHub, Google, Stripe, Slack), passwords, tokens, private keys, JWTs, database URLs with credentials. Runs a fast regex scan, then triages matches to remove false positives and checks .env/git hygiene. Use when auditing a codebase for leaked credentials or before committing/publishing a repo.
license: MIT
---

# Secrets Detection

## Step 1 — Fast regex scan

```bash
bash scripts/scan.sh <project_root>
```

The script greps the pattern set in `references/patterns.txt` (one ERE regex per line — edit that file to tune, not the script). Output: `file:line:matched line`.

If `grep -rE` is unavailable, replicate manually with `rg -n` using the same patterns file.

## Step 2 — Triage (kill false positives)

For each hit, read the surrounding code and drop it if the value is:

- A placeholder: `example`, `test`, `dummy`, `sample`, `changeme`, `placeholder`, `xxx`, `your_...`, `todo`, `<...>`, `***`, `...`
- In an obvious fixture/mock: paths containing `test/`, `__tests__/`, `fixtures/`, `examples/`, `docs/`, `*.md` code blocks
- A variable reference, not a literal: `API_KEY=$API_KEY`, `password: process.env.PW`
- An intentional canary/honeypot token (look for comments like `canary`, `honeypot`, or a monitoring URL)

Keep — and flag as CRITICAL — anything that looks real: correct prefix + realistic entropy + committed to git history.

## Step 3 — Hygiene checks

```bash
git ls-files | rg -i '\.env($|\.)'          # .env files tracked?
cat .gitignore 2>/dev/null | rg -i 'env|secret|key|pem'   # .env ignored?
git log --all --oneline -S 'AKIA' -- '*.py' 2>/dev/null | head  # leaked keys in history?
rg -n --hidden -g '!*.lock' -g '!package-lock.json' '(BEGIN (RSA |EC |OPENSSH )?PRIVATE KEY)' <root>
```

Report:

1. **Committed real secrets** → CRITICAL. Fix: rotate/revoke immediately (a removed file is not a revoked key), then scrub history (`git filter-repo` / BFG).
2. **Tracked `.env` files** → CRITICAL regardless of content.
3. **Secrets in git history but not HEAD** → HIGH (history is forever on clones/forks).
4. **Placeholder secrets in prod-looking config** → MEDIUM (someone will "temporarily" deploy them).
5. **Secrets handling patterns** — note whether the project uses env vars / secret manager properly, and whether secrets can leak into logs (see also `../data-exposure/SKILL.md`).

## Step 4 — Dev-artifact leak surface (developer tools commit secrets too)

```bash
rg --files --hidden -g '.vscode/**' -g '.idea/**' -g '*.postman_collection.json' -g '*.postman_environment.json' -g '*.http' -g '*.rest' -g 'devcontainer.json' -g '*.insomnia*' | head
rg -n -i "token|secret|password|api[_-]?key|authorization" .vscode/ .idea/ *.postman* *.http devcontainer.json 2>/dev/null | head
```

- **Postman/Insomnia collections & environments**: live `Authorization: Bearer …` headers, saved logins — these files are SHARED and committed by design → CRITICAL (rotate on sight)
- **`.vscode/launch.json` / `.idea/workspace.xml`**: `env`/`ENV` blocks with tokens and DB URLs → HIGH; also debug `console`/`attach` configs revealing internal hosts
- **`.http`/`.rest` request files**: same class — request headers with real credentials → HIGH/Critical
- **`devcontainer.json` / `codespaces`**: secrets in `containerEnv`/`remoteEnv`, PATs in feature options → HIGH
- **Swagger/OpenAPI `examples:`** that are real payloads (real users, real tokens) → Medium/High

Same triage rules apply (fakes/fixtures dropped); these surfaces pair with the generic patterns from Step 1 — the difference is WHERE to look, not WHAT to match.

## Notes

- Scan binary-free (`-I`) and exclude vendored dirs — the script does this already.
- Encoded secrets (base64, hex blobs near words like `key`, `token`, `secret`) deserve a look: decode and judge.
- `references/patterns.md` documents each pattern, its false-positive profile, and how to add new ones.
- **Optional bridge:** if `gitleaks`/`trufflehog` is installed (probe: `../security-audit/scripts/probe_tools.sh`), run it too — it scans **full git history**, where worktree regex cannot see deleted-then-rotated keys. Merge hits into Step 2 triage; same false-positive rules apply.

Files in this skill

  • LICENSE1 KB
  • SKILL.md3 KB
  • SOURCE.md374 B
  • TRUST.auto.yaml2.1 KB
  • references/patterns.md2.5 KB
  • references/patterns.txt2.2 KB
  • scripts/scan.sh931 B

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…