The Agent Skills Directory

[COMMAND_EXECUTION]: The skill performs command execution by calling local Node.js scripts located in the .specforge/core/scripts/ directory. These tools, including artifact-graph-status.mjs, quality-suite.mjs, and gate.mjs, are used to analyze project state and update readiness gates based on review findings.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes content from untrusted project files like requirements.md and technical-design.md. While no active exploit was found, this represents a potential attack surface.
Ingestion points: Reads local specification files such as brief.md, prd.md, requirements.md, ui-design.md, technical-design.md, and tasks.md.
Boundary markers: No explicit delimiters or instructions are provided to the agent to treat the content of these files as untrusted data.
Capability inventory: The skill can execute local project-related scripts and modify the state of project approval gates.
Sanitization: There is no evidence of sanitization or filtering applied to the input specification files before they are processed by the agent.

sf-spec-review