# Log Analyzer

> |

- **Type:** Skill
- **Install:** `agentstack add skill-holsee-conjure-log-analyzer`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [holsee](https://agentstack.voostack.com/s/holsee)
- **Installs:** 0
- **Category:** [Agent Skills](https://agentstack.voostack.com/c/agent-skills)
- **Latest version:** 0.1.0
- **License:** Apache-2.0
- **Upstream author:** [holsee](https://github.com/holsee)
- **Source:** https://github.com/holsee/conjure/tree/main/priv/skills/log-analyzer

## Install

```sh
agentstack add skill-holsee-conjure-log-analyzer
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

# Log Analyzer Skill

A production monitoring skill for fetching, analyzing, and diagnosing issues from application logs.

## Available Scripts

### 1. Fetch Logs (`fetch_logs.py`)

Fetches logs from a REST API endpoint.

```bash
# Fetch last 100 logs
python3 scripts/fetch_logs.py --endpoint "http://localhost:8080/api/logs" --limit 100

# Fetch logs with time range
python3 scripts/fetch_logs.py --endpoint "http://localhost:8080/api/logs" \
  --start "2024-01-15T00:00:00" --end "2024-01-15T23:59:59"

# Fetch logs by severity
python3 scripts/fetch_logs.py --endpoint "http://localhost:8080/api/logs" \
  --level ERROR
```

**Options:**
- `--endpoint` - REST API URL for logs (required)
- `--limit` - Maximum number of logs to fetch (default: 100)
- `--start` - Start time (ISO 8601 format)
- `--end` - End time (ISO 8601 format)
- `--level` - Filter by log level (DEBUG, INFO, WARN, ERROR)
- `--output` - Output file (default: stdout)

### 2. Parse Logs (`parse_logs.py`)

Parses log files in various formats.

```bash
# Parse JSON logs
python3 scripts/parse_logs.py --format json logs.json

# Parse text logs with pattern
python3 scripts/parse_logs.py --format text --pattern "%(timestamp)s %(level)s %(message)s" app.log

# Parse and filter by level
python3 scripts/parse_logs.py logs.json --level ERROR
```

**Supported Formats:**
- `json` - JSON lines format
- `text` - Plain text with pattern
- `auto` - Auto-detect format

### 3. Analyze Logs (`analyze.py`)

Analyzes logs for patterns, errors, and anomalies.

```bash
# Full analysis
python3 scripts/analyze.py logs.json

# Error analysis only
python3 scripts/analyze.py logs.json --errors-only

# Generate summary
python3 scripts/analyze.py logs.json --summary

# Find patterns
python3 scripts/analyze.py logs.json --patterns
```

**Analysis Types:**
- Error frequency and patterns
- Response time anomalies
- Request volume trends
- Common error messages
- Suggested diagnostics

## Example Workflow

1. **Fetch recent logs:**
   ```bash
   python3 scripts/fetch_logs.py --endpoint "http://api.example.com/logs" \
     --limit 500 --output recent_logs.json
   ```

2. **Analyze for errors:**
   ```bash
   python3 scripts/analyze.py recent_logs.json --errors-only
   ```

3. **Get summary:**
   ```bash
   python3 scripts/analyze.py recent_logs.json --summary
   ```

## Log Format Reference

See `references/log_formats.md` for supported log formats and examples.

## Tips

- Always start by fetching recent logs with a reasonable limit
- Use `--errors-only` for quick triage
- Use `--summary` to understand overall health
- Check `--patterns` to find recurring issues

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [holsee](https://github.com/holsee)
- **Source:** [holsee/conjure](https://github.com/holsee/conjure)
- **License:** Apache-2.0

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** yes
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** no
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-holsee-conjure-log-analyzer
- Seller: https://agentstack.voostack.com/s/holsee
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.