skills/gaoscode/planweave/plan-maker/Gen Agent Trust Hub

plan-maker

Pass

Audited by Gen Agent Trust Hub on Jul 16, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill's primary function is to provide structured instructions for drafting project plans. It does not contain any malicious commands, obfuscated content, or unauthorized data access patterns.
  • [COMMAND_EXECUTION]: The skill instructions mention interacting with a 'planweave' CLI for workspace operations and file materialization. These operations are within the stated purpose of the skill and are performed when the user explicitly requests to materialize or execute a plan.
  • [PROMPT_INJECTION]: The skill processes untrusted user input and codebase content to generate plan drafts, which represents a surface for indirect prompt injection. The skill defines clear boundaries for when the agent should transition from drafting to execution, reducing the likelihood of accidental execution of injected instructions.
  • Ingestion points: User goals, requirements, and local codebase context including README files, source code, and schemas.
  • Boundary markers: Absent; the skill does not specify delimiters for external input in its prompt construction logic.
  • Capability inventory: File writing (project-graph.json, manifests, prompt files) and execution of the 'planweave' CLI.
  • Sanitization: Absent; no explicit sanitization or validation of input data is described in the instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 16, 2026, 05:27 AM
Security Audit — agent-trust-hub — plan-maker