serena-conventions

Pass

Audited by Gen Agent Trust Hub on Aug 24, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill facilitates reading source code and performing structural modifications, creating a vector for indirect prompt injection where malicious instructions embedded in project files could influence agent behavior. \n
  • Ingestion points: Project files accessed via serena_find_symbol, serena_get_symbols_overview, and ctx_batch_execute cat in SKILL.md. \n
  • Boundary markers: Absent; no specific instructions are provided to the agent to delimit or ignore embedded commands in read content. \n
  • Capability inventory: Includes tools for file modification (serena_replace_symbol_body, serena_insert_after_symbol) and shell command execution (bash). \n
  • Sanitization: Absent; no content validation or escaping is specified for ingested code.
  • [COMMAND_EXECUTION]: The skill utilizes shell commands (bash) and sandbox execution (ctx_execute) for file system operations, git tasks, and running developer tools like Pint. These capabilities are consistent with the skill's intended purpose in a software development environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 24, 2026, 04:13 AM
Security Audit — agent-trust-hub — serena-conventions