The Agent Skills Directory

[SAFE]: The skill implements a structured workflow for assessing business value. It uses specific templates and policies (references/research-policy.md) to guide agent behavior during data collection.
[SAFE]: No evidence of prompt injection or attempts to bypass safety guidelines. The skill specifically includes instructions to separate facts, inferences, and assumptions, which improves the reliability of the output.
[SAFE]: Network operations are limited to 'Web Search' as part of the research stage. These are governed by a research plan and results are logged to sources/sources.json for auditability.
[SAFE]: File access is restricted to the local project workspace (reading upstream stage documents and writing to specific output files like 05-value.md). No access to sensitive system directories or credentials was found.
[SAFE]: The disable-model-invocation: true flag in the YAML frontmatter is a defensive configuration that limits the agent's execution environment to the instructions provided.

value-assess