Works with the AI Agents you already use
Semantic Entropy Credential Scanner
Perform deep forensic scans using Shannon entropy and semantic analysis to find obfuscated secrets and leaked keys.
$15
Semantic Entropy Credential Scanner
Example session with this skill installed
Run a semantic entropy scan on the current repository to find any obfuscated keys or high-entropy tokens in the codebase.
- Read your context and instructions
- Compiled the semantic entropy credential
I have completed the semantic and statistical entropy scan of the workspace. I analyzed 42 files, focusing on string literals and variable declarations.
Credential Forensics Audit Report
| File Path | Variable/Context | Entropy Score | Risk Level | Finding Type |
|---|---|---|---|---|
src/auth/provider.ts | const _k = "S0m3R4nd0m..." | 7.82 | CRITICAL | High-entropy string literal |
config/settings.json | "api_key": "base64:ZXhw..." | 6.94 | HIGH | Base64 encoded payload |
scripts/deploy.sh | AUTH_TOKEN | 5.12 | MEDIUM | Semantic match (Environment Var) |
.env.local | STRIPE_SK | 7.45 | CRITICAL | High-entropy secret |
tests/mocks.js | test_uuid | 3.21 | LOW | False Positive (Standard UUID) |
Analysis Summary
- Total Strings Scanned: 1,248
- High Entropy Hits (>6.5): 3
- Semantic Matches: 2
- Action Required: 3 critical/high findings suggest active secrets are present in the source.
Next steps
- Move the identified keys in
src/auth/provider.tsand.env.localto a secure vault or managed secret provider. - Run
git filter-repoor a similar tool if these secrets were previously committed to version control. - Update
config/settings.jsonto use environment variable references instead of encoded strings.
Connects securely to your tools. The creator never sees your data.
About this skill
Manual pattern matching misses the cleverest obfuscation. Standard linters catch hardcoded strings, but they fail when secrets are hidden in base64 blobs, high-entropy variable assignments, or deep git history. This skill implements advanced credential forensics to find what static analysis ignores.
What it does
- Shannon Entropy Analysis flags non-random cryptographic tokens by calculating statistical randomness in string literals.
- Semantic Context Probing evaluates variable naming conventions like
token,auth, andhashto identify suspicious declarations. - Obfuscation Detection uncovers base64-encoded payloads and disguised keys hidden in configuration files.
- Prioritized Reporting delivers a filtered list of findings, distinguishing between standard UUIDs and actual leaking credentials.
How it works
- Scope Scan initiates a recursive analysis of string literals and variable declarations across the workspace.
- Entropy Calculation runs a Shannon entropy pass to identify high-density tokens that deviate from standard human text.
- Context Filtering cross-references high-entropy strings with surrounding code clues and naming patterns.
- Result Verification eliminates false positives from standard hashes or public identifiers before presenting the final audit.
Frameworks & tools
This skill operates as a language-agnostic forensic tool suitable for any codebase. It utilizes local environment runners like npx, uvx, or pipx to maintain workspace integrity and ensures no data leaves the local memory.
Why this beats prompting it yourself
Generic prompts often hallucinate security risks or miss encoded strings entirely. This skill follows a deterministic forensic workflow that combines statistical math with semantic awareness, ensuring deep coverage without the noise of false positives.
Use cases
- Audit legacy codebases for hardcoded production secrets before a public release.
- Scan git commit blobs for obfuscated keys buried in historical data.
- Validate configuration files to ensure no base64-encoded credentials are present.
Known limitations
Analysis is restricted to the active project workspace. It will not scan system-level configuration paths or parent directories.
Restricted to the active project workspace. Cannot access parent directories or global system paths. Analysis is performed in-memory only.
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 any payment.
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
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 4 days ago
Requires
Permissions
Target Use Cases: - Scan for obfuscated or base64-encoded credentials in my code. - Find high-entropy hidden secrets in git history. - Audit source code for disguised API keys.