Log Analyzer

PassAudited by ClawScan on May 1, 2026.

Overview

This is a coherent local log-analysis tool with no network, install, credential, or persistence behavior, though its reports can include raw log snippets.

This appears safe to install for local log analysis. Before using it, make sure the selected log files do not contain secrets or private data you do not want included in the agent conversation or exported reports.

Findings (1)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

Low

#ASI06: Memory and Context Poisoning

What this means

Sensitive strings present in logs could be surfaced in the agent chat, JSON output, or reports.

Why it was flagged

The analyzer includes raw portions of log lines as examples in its output. This is purpose-aligned, but log lines may contain secrets, personal data, or untrusted text that should be treated as data rather than instructions.

Skill content

error_examples[fp] = stripped[:500]
...
"example": error_examples.get(fp, "")

Recommendation

Run it only on intended log files, redact secrets when needed, and treat any displayed log content as untrusted data.