The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external, untrusted data including error messages, stack traces, and log entries during the reproduction and isolation phases.
Ingestion points: Data enters the agent's context through triggers like 'error', 'traceback', and 'log analysis' as described in SKILL.md and references/systematic-debugging.md.
Boundary markers: The instructions lack explicit delimiters or safety warnings for the agent to ignore potentially malicious instructions embedded within the processed logs or error strings.
Capability inventory: The skill environment permits the execution of powerful command-line tools (pdb, node, delve), which could be targets for manipulation via injected instructions.
Sanitization: No logic is present to sanitize or escape data extracted from external logs before it is processed by the agent.
[DYNAMIC_EXECUTION]: The skill provides the agent with instructions and templates for executing powerful debugging and version control tools.
Evidence: SKILL.md and references/debugging-tools.md include commands for launching language-specific debuggers such as python -m pdb, node --inspect-brk, and dlv debug, as well as performing regression hunting with git bisect. While intended for legitimate debugging, these tools grant the agent significant control over the local execution environment.
[METADATA_POISONING]: There is an inconsistency between the skill's internal author metadata (https://github.com/Jeffallan) and the provided vendor context.
Evidence: The SKILL.md file identifies the author as https://github.com/Jeffallan, which differs from the provided context. While this may be a fork, it represents a metadata discrepancy in the skill's identity.

debugging-wizard