The Agent Skills Directory

[SAFE]: No malicious patterns or security risks were identified. The skill defines a structured workflow for plan creation and uses platform-provided MCP tools for persistence.
[PROMPT_INJECTION]: The skill processes untrusted data from task descriptions and existing artifacts, creating a potential surface for indirect prompt injection. However, the risk is minimal as the skill lacks high-privilege capabilities and explicitly forbids modifying source code.
Ingestion points: Task descriptions, linked context, and artifact content fetched via mcp__aiden__get_artifact in SKILL.md.
Boundary markers: Absent.
Capability inventory: Artifact management tools (mcp__aiden__create_plan, mcp__aiden__update_artifact, mcp__aiden__add_artifact_comment, mcp__aiden__resolve_artifact_comment) and read-only codebase exploration (Read, Grep, Glob). File writing/modification is explicitly forbidden.
Sanitization: Absent.

aiden-plan