The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The installation instructions include npx clawhub install factor-backtester. This command executes a package from a third-party source that is not a trusted organization or well-known service, creating a potential supply chain risk.
[COMMAND_EXECUTION]: The script scripts/backtest_factor.py uses pathlib.Path(args.input_path).read_text() to read files. This creates a file-read capability that could be misused to access sensitive local files if the agent is directed to provide malicious paths as arguments to the script.
[INDIRECT_PROMPT_INJECTION]: The skill processes external JSON data files which are then parsed and summarized for the agent. This represents an ingestion surface for untrusted data that could influence agent behavior if the source data is compromised.
Ingestion points: scripts/backtest_factor.py reads data from files such as assets/factor_alpha_v1.json via command-line arguments.
Boundary markers: The script lacks explicit delimiters or instructions to the agent to ignore potentially embedded instructions in the processed data.
Capability inventory: The script performs file-read operations and prints structured JSON summaries to the console.
Sanitization: No validation or sanitization is performed on the file content beyond standard JSON parsing before the data is processed and returned to the agent context.

factor-backtester