The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests and acts upon instructions contained within YAML templates (e.g., prd-template, custom-epic-tmpl.yaml), creating a surface for indirect prompt injection if those templates contain adversarial content. \n
Ingestion points: Templates are loaded from the local resources/ directory or file paths provided by the user as described in SKILL.md. \n
Boundary markers: The skill enforces a mandatory 1-9 elicitation format and detailed rationale blocks to separate the agent's drafted content and reasoning from the structured data provided in the template. \n
Capability inventory: Capabilities include writing incrementally to files, modifying markdown documents, and calling other skills like mermaid-architect and pm-checklist. \n
Sanitization: Content drafting is guided by template instructions, and the skill requires a hard stop for user validation in any section marked with elicit: true. \n- [COMMAND_EXECUTION]: Documentation included with the skill (references/document-change-log.md and references/document-status-lifecycle.md) provides shell command snippets, specifically using node -e, to programmatically update document change logs. \n
Evidence: command node -e 'const fs = require("fs"); const CL = require("./.agents/skills/{develop|finalise}/references/change-log.js"); ...' "$DOC" ... \n
Context: These commands are localized to the agent's skill environment and reference internal project logic for maintaining document consistency and history.

create-doc