Skip to content

[P2][security] Data exposure: published identifiers, path traversal in start-server.sh, placeholder tokens in docs #9

Description

@Toqsick

Source: RepoLens Security-Audit 2026-08-16 (findings security/data-exposure/004–011, input-sanitization/003, cryptography/005, adversarial-verified ✅)

  • Published personal/infra identifiers in skill metadata and references: Telegram chat + supergroup IDs, Hermes host/network audit content, home-network/Wi-Fi fingerprint, owner + cloud resource identifiers
  • input-san/003 start-server.sh --project-dir concatenated into SESSION_DIR without canonicalization (.. traversal)
  • data-exposure/012 hardcoded dashboard session tokens in docs (my-static-token, test-token-hermes-mobile-2026)
  • cryptography/005 MD5 for non-security IDs (fine, but document the intent)
  • LOW: raw exception text in 500s, full tracebacks in PDF scripts, Harbor default password echo

Suggested fix: scrub identifiers from docs, realpath the project-dir, replace placeholder tokens with <your-token> style, central exception handler.

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions