session-cleaner

What This Skill Does

The session-cleaner skill is a specialized utility designed to transform raw, machine-heavy JSONL session logs generated by OpenClaw or Clawdbot into human-readable markdown transcripts. When AI agents interact, the underlying data structures are cluttered with verbose tool-use logs, system-level metadata, and raw JSON artifacts that are difficult to parse manually. This skill acts as a parser and formatter, effectively stripping away the technical noise while preserving the logical flow of the conversation between the user and the assistant. By prioritizing the narrative content, it makes session auditing, documentation, and knowledge sharing significantly more efficient for team collaboration.

Installation

To integrate this skill into your environment, utilize the OpenClaw repository manager. Run the following command in your terminal: clawhub install openclaw/skills/skills/henrino3/ec-session-cleaner Ensure you have Node.js installed, as the primary script session-cleaner.mjs requires a runtime environment to execute the parsing logic effectively. For advanced users managing multiple agent logs, verify that your directory permissions allow for local script execution and file modifications.

Use Cases

This skill is indispensable for developers and project managers who perform quality assurance on agent interactions. Primary use cases include: 1. Post-mortem analysis of agent errors to see what was actually communicated versus what was calculated. 2. Generating clean documentation for stakeholders to review project progress without overwhelming them with raw data. 3. Training data curation by archiving successful conversation histories into standardized formats. 4. Streamlining remote agent maintenance using the included shell scripts for bulk processing across distributed servers.

Example Prompts

1. "Clean the session log from yesterday's Spock agent interaction and save it to the documentation folder."
2. "Run the batch cleaner on all files in the logs directory and output the markdown results into my summary folder."
3. "Process the remote Scotty logs via the SSH cleaner script and show me the final human-readable transcript."

Tips & Limitations

When using the cleaner, ensure your JSONL files are well-formed; corrupted JSON strings may cause the parser to fail. The skill currently focuses on standard OpenClaw formats, so non-standard log schemas might not be fully scrubbed of metadata. For high-volume environments, use the session-cleaner-spock.sh script to minimize system overhead by processing logs in batches rather than calling the primary script individually. Always back up your original raw logs before running batch operations to prevent accidental data loss.