The Agent Skills Directory

[COMMAND_EXECUTION]: The SKILL.md file contains a 'Memory Protocol' section that explicitly instructs the agent to execute a shell command (cat .claude/context/memory/learnings.md).
Evidence: The instruction block is formatted as a bash command intended to be run by the agent to retrieve previous context.
Risk: This establishes a pattern of direct command execution. If an attacker can influence the file path or the environment, this capability could be misused.
[PROMPT_INJECTION]: The skill implements an automated context-loading mechanism that is susceptible to indirect prompt injection.
Ingestion points: The agent is instructed to read content from .claude/context/memory/learnings.md at the start of every session.
Boundary markers: Absent. The skill does not define delimiters or provide 'ignore embedded instructions' warnings for the data loaded from the memory file.
Capability inventory: The agent has access to Read, Write, and Edit tools, as well as the ability to execute cat as described in the memory protocol.
Sanitization: Absent. There is no validation or filtering performed on the content retrieved from the memory file before it is integrated into the agent's current context.

fiber-logging-and-project-structure