The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external project files which could potentially contain adversarial instructions designed to influence the agent's behavior.
Ingestion points: The skill reads content from docs/01-需求调研/需求调研报告.md and docs/01-需求调研/MVP方案.md (SKILL.md).
Boundary markers: There are no explicit delimiters or instructions to the model to ignore embedded prompts within the source documentation.
Capability inventory: The skill utilizes Write permissions to create files in the docs/ directory and has access to the Bash tool for environment checks (SKILL.md).
Sanitization: The skill does not implement specific sanitization or validation logic for the text read from markdown files before processing it.
[COMMAND_EXECUTION]: The skill executes shell commands via the Bash tool to perform pre-execution environment checks.
Evidence: The 'Preamble' section contains a bash script that uses [ ! -f ... ] to verify the presence of required documentation before proceeding (SKILL.md).

pm-journey