The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests and processes data from external sources to generate summaries and logs.
Ingestion points: Data is retrieved from Azure DevOps work item descriptions (wit_get_work_item), Git commit messages, and file contents during the backfilling process.
Boundary markers: The skill does not define explicit delimiters or instructions to the agent to ignore potentially malicious embedded content within the work items or commits.
Capability inventory: The skill has the ability to write to Azure DevOps (wit_add_work_item_comment, wit_update_work_item) and publish to documentation backends like Notion or Confluence.
Sanitization: There is no evidence of sanitization or validation of the content retrieved from the work items or git history before it is re-processed or mirrored.
[EXTERNAL_DOWNLOADS]: The skill references a external repository (https://github.com/Pimzino/spec-workflow-mcp) as the source for the underlying implementation log schema.
[COMMAND_EXECUTION]: The skill performs local shell operations during its 'Backfilling Logs' workflow.
Evidence: It uses git log --follow, git show, and grep to extract implementation details from the repository history when logs are missing.

ado-impl-logger