The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a Python script using shell-like interpolation of user-provided inputs, such as competitor names and descriptions of their moves, which could be exploited if inputs are not properly sanitized.
Evidence: The process step 4 calls python "${CLAUDE_PLUGIN_ROOT}/scripts/narrative-mapper.py" --brand {slug} --action generate-counter --competitor {name} --their-move "{what they did}" --move-type {type} using variables provided by the user.
[DYNAMIC_EXECUTION]: The skill relies on the execution of a local Python script (narrative-mapper.py) located within the plugin directory to perform its core strategy generation logic.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external, untrusted competitor data which could contain malicious instructions meant to influence the agent's analysis or output.
Ingestion points: The skill accepts "Evidence" in the form of URLs to competitor messaging, social posts, ads, and press releases.
Boundary markers: No explicit delimiters or instructions to ignore embedded commands within the analyzed evidence are defined in the process.
Capability inventory: The skill has the ability to read local brand configuration files (~/.claude-marketing/), execute scripts via the shell, and coordinate multiple specialized agents (competitive-intel, content-creator, marketing-strategist).
Sanitization: There is no evidence of sanitization or filtering of the external content before it is processed by the analysis agents.

counter-narrative