make-tweakable

Pass

Audited by Gen Agent Trust Hub on May 2, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill establishes an indirect prompt injection surface by creating a feedback loop between user-controlled browser state and agent actions.\n
  • Ingestion Point: The pending.yaml file (referenced in Phase 4 and Step 6) is intended to be read by the agent after being written by a client-side script.\n
  • Boundary Markers: The skill does not define specific delimiters or instructions for the agent to ignore embedded commands within the YAML values.\n
  • Capability Inventory: The agent is instructed to use the Edit tool to modify local HTML files based on the contents of the YAML file.\n
  • Sanitization: No sanitization or schema validation is defined for the values received from the client-side script before they are processed by the agent.\n- [COMMAND_EXECUTION]: The skill executes shell commands to set up its working environment.\n
  • Evidence: Uses the Bash tool with mkdir -p in Phase 4 to prepare directory structures for session state and artifact storage.
Audit Metadata
Risk Level
SAFE
Analyzed
May 2, 2026, 02:20 PM
Security Audit — agent-trust-hub — make-tweakable