The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or unauthorized data access were detected. The skill interacts with the vendor's command-line interface as intended for task history management.
[PROMPT_INJECTION]: The skill contains a vulnerability surface where historical data from past runs is ingested into the current agent session. This is an inherent risk of memory-retrieval features.
Ingestion points: Results from parallel-cli memory retrieve (such as input_excerpt and output_excerpt) are used throughout SKILL.md to inform agent context.
Boundary markers: The skill instructs the agent to "Distinguish recalled information from any fresh verification," which acts as a behavioral boundary, although no structural delimiters (like XML tags) are specified for the tool output.
Capability inventory: The skill is restricted to the parallel-cli toolset via the allowed-tools configuration.
Sanitization: No explicit sanitization or filtering of historical data is performed within the skill instructions, as the data is expected to be managed by the underlying CLI tool.

parallel-memory