The Agent Skills Directory

[SAFE]: The skill follows security best practices by implementing explicit 'Hard Stops' and 'Human Gates' in SKILL.md and references/human-gates.md. these rules strictly prohibit the agent from performing sensitive actions such as publication, external communication, or system modifications without human approval.
[SAFE]: Indirect Prompt Injection surface analysis: (1) Ingestion points: The skill reads local project state files as defined in references/state-sources.md (e.g., ACTIVE_STATE.md, ROADMAP.md). (2) Boundary markers: No explicit delimiters or 'ignore' instructions are used for the ingested content. (3) Capability inventory: The skill instructions in SKILL.md and references/human-gates.md confirm the agent has no access to network tools, file-writing capabilities, or shell execution for this workflow. (4) Sanitization: No sanitization is performed on ingested project data. Because the skill's only output is a structured text packet, the risk associated with this ingestion surface is negligible.

themindshift-loop-router