The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill provides a CLI alternative command using npx @claude-flow/cli@latest. This instructs the system to download the latest version of the @claude-flow/cli package from the NPM registry.
[COMMAND_EXECUTION]: The skill requests access to the Bash tool and provides shell commands for data management tasks.
[DYNAMIC_EXECUTION]: The use of npx to fetch and execute code dynamically at runtime is a potential vector for supply chain attacks, especially when using the @latest tag which bypasses version pinning.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest data from untrusted external sources like REST APIs and CSV files.
Ingestion points: Step 1 in SKILL.md explicitly mentions fetching data from REST APIs and CSV files.
Boundary markers: There are no explicit boundary markers or instructions to the agent to ignore embedded commands within the ingested data.
Capability inventory: The skill utilizes Bash, memory storage tools (mcp__plugin_ruflo-core_ruflo__memory_store, mcp__plugin_ruflo-core_ruflo__memory_search), and HNSW indexing tools (mcp__plugin_ruflo-core_ruflo__ruvllm_hnsw_create, mcp__plugin_ruflo-core_ruflo__ruvllm_hnsw_add).
Sanitization: While the skill performs mathematical normalization on price and volume data (which acts as a form of sanitization for numerical values), it lacks validation for non-numerical content that may be included in the market data source.

market-ingest