The Agent Skills Directory

[COMMAND_EXECUTION]: The skill contains instructions for the agent to execute shell commands and local scripts to manage its operational state.
Evidence includes instructions to run scripts/stelow next and scripts/stelow advance <next-candidate> to handle workflow transitions.
The skill also includes a bash snippet to check for the presence of STELOW_WORKFLOW and STELOW_STATE environment variables when the skill is loaded to determine its execution mode (standalone vs. workflow).
These commands are functional components of the author's 'stelow' framework and are used for workflow coordination.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it ingests and processes untrusted user data related to product strategy.
Ingestion points: User-provided strategy queries, roadmap descriptions, and experimentation inputs defined in the interaction examples in SKILL.md.
Boundary markers: Absent; there are no specific instructions or delimiters provided to separate user-supplied content from the skill's core strategic instructions.
Capability inventory: The agent is authorized to execute local scripts (scripts/stelow) and perform shell environment checks.
Sanitization: No explicit sanitization or validation logic is defined for the incoming user input before it is processed by the strategic reflection lenses.

stelow-product-evolutionary-principles