Scan oversized logs with Gemini-backed grep and summarization modes.
npx skills add https://github.com/marin-community/marin --skill scan-logs
Please help me install this Agent Skill. Skill: scan-logs Source: https://github.com/marin-community/marin/tree/main/.agents/skills/scan-logs Command: npx skills add https://github.com/marin-community/marin --skill scan-logs
Oversized log files are too large to read manually or paste into a chat context, making it hard to locate errors or understand failures. This Skill runs a script that searches and summarizes large logs using the Gemini API.
Ask the assistant to scan the large log file for error patterns and produce a summarized Markdown report using the scan-logs script.
High-intent search queries and answers about installing and using this skill.
Run the logscan script in grep mode with the log file path and your query, for example: uv run scripts/logscan.py grep log.txt "errors". It returns line-numbered original lines matching the query, with progress reported to stderr.
Use summarize mode, which hierarchically reduces large inputs in 50,000-token chunks and outputs a Markdown report to stdout. You can also pipe grep results into summarize with --stdin to narrow the input first.
Yes, GEMINI_API_KEY must be set because the log content is sent to the external Gemini API. This incurs model usage costs, so avoid sending logs containing secrets or sensitive data without approval.
Yes, use --chunk-tokens, --concurrency, and --model options. Defaults are 5,000-token grep chunks, 50,000-token summarize chunks, concurrency 16, and gemini-2.5-flash-lite; check --help before overriding.
Avoid it when logs contain secrets or sensitive data that cannot leave your environment, since content is sent to an external API. Also verify generated diagnoses against the matched source lines before acting on them.
Cogni-DAOCollect and analyze logs from multiple environments using LogQL queries.
kwam1naTrace bug causal chains through phased debugging workflows with evidence collection.
profmikegreeneDetect emerging anomalies and risk flags in live pilot metrics and data streams.
yun520-1Parse and search application logs using grep, awk, jq, and Python3.
