symph0nia/cyberedge/cyberedge-monitor-public-code

Claude Code and agent skill by symph0nia/cyberedge

Trust tier 3
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

Search bounded GitHub public-code metadata for exact domain names in an explicitly authorized CyberEdge Scope. Use when an AI operator must run public-code intelligence, inspect evidence-backed reference candidates, or distinguish completed negative coverage from provider failure without retrieving source content or claiming confirmed credential leakage.

Install

npx vskill@latest install symph0nia/cyberedge/cyberedge-monitor-public-code

SKILL.md

CyberEdge Monitor Public Code

Use cyberedge-agent; this machine bridge is not a human CLI.

Preconditions

  • Require an existing Scope with a non-empty human authorization reference and at least one domain target.
  • Require this exact Skill/version grant with scan.intelligence, task.read, finding.read, evidence.read, and report.read.
  • Require the isolated provider adapter. The GitHub token belongs only to that adapter.

Workflow

  1. Start policy_public_code_intelligence for the authorized Scope.
  2. Watch the Task until it becomes terminal.
  3. Correlate each detector=github-public-code Finding with its github.code.reference Observation and JSON Evidence.
  4. Report the queried domain, repository, path, blob SHA, GitHub URL, Evidence ID, and lifecycle state.
  5. Treat github.code.coverage as a completed negative evaluation for that exact domain.
  6. Treat github.code.error or github.code.no_targets as unknown coverage. Do not resolve or suppress prior candidates from an error.

Every result is an Info review candidate. It proves only that GitHub code search returned public repository metadata referencing the domain. It does not prove that a secret, credential, exploit, or sensitive source is present.

Never fetch repository content, clone a returned repository, print or request the provider token, widen the query, search domains outside the Scope, or describe a candidate as a confirmed leak.

json
{"request_id":"req-1","idempotency_key":"idem-1","agent_id":"codex-main","skill_name":"cyberedge-monitor-public-code","skill_version":"0.1.0","action":"start_scan","scope_id":"scope-id","policy_id":"policy_public_code_intelligence"}

Security

  • Security scanPassed, 90 of 100Pattern checks on the skill's files.
  • Model reviewPassed

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

Repository stars
611
Repository forks
75
7-day trend score
32

Stars and forks are for the whole GitHub repository, not this skill alone. Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/symph0nia/cyberedge/cyberedge-monitor-public-code/badge)](https://verified-skill.com/skills/symph0nia/cyberedge/cyberedge-monitor-public-code)

Something wrong with this skill or its page? Tell us.