The Agent Skills Directory

[SAFE]: The skill performs legitimate documentation management tasks within the context of a local repository. It does not utilize network tools, request sensitive credentials, or attempt to execute shell scripts.\n- [PROMPT_INJECTION]: The skill has an indirect prompt injection surface because it reads untrusted data from the repository (existing ADRs, source code, and design documents) to inform its output. However, the impact is limited to the generation of documentation prose, and no dangerous tools or autonomous execution paths are exposed to this input.\n- [DATA_EXPOSURE]: The skill accesses repository content, including code paths, interfaces, and internal documentation. This access is necessary for the skill's primary purpose and is restricted to the local environment without any exfiltration patterns detected.

adr-manager