The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes a specialized CLI tool, effect-solutions, to fetch the latest recommendations for various topics such as tsconfig, error-handling, and testing. This allows the agent to provide advice based on external, up-to-date standards rather than relying on its training data.
[SAFE]: The skill includes comprehensive documentation on managing sensitive information. It explicitly instructs developers to use Config.redacted and the Redacted type for secrets like encryption keys, ensuring they are not exposed in logs or traces.
[SAFE]: Guidance on framework integration includes the use of well-known and trusted services like Sentry for observability and tracing, following industry standards for application monitoring.
[PROMPT_INJECTION]: No patterns of behavior override or safety bypass were detected. The instructions are focused on technical guidance and project-specific rules.
[DATA_EXFILTRATION]: There are no indicators of unauthorized data transfer. The skill's network behavior is limited to the implied documentation fetching via the documented CLI tool.
[INDIRECT_PROMPT_INJECTION]: The skill exhibits an ingestion surface where external content (CLI output) is introduced into the agent's context. This data is used to inform code generation and review.
Ingestion points: Output from the effect-solutions command in SKILL.md.
Boundary markers: None explicitly defined for the CLI output.
Capability inventory: File modification in packages/services/ and agent-tools/* as described in SKILL.md.
Sanitization: None specified for the documentation content.

effect-ts