llm-summarize

Structured session insight extraction for knowledge systems.

Philosophy

- Config-driven - No hardcoded defaults, specify exact provider/model
- Prismis pattern - Secrets in .env, references in config.toml via env:VAR_NAME
- Knowledge-focused - Extracts decisions, patterns, preferences, not just summaries
- Composable - JSON output pipes to jq and other tools

Installation

``bash
cd llmcli-tools
./install.sh llm-summarize
`

Configuration

$3

`toml
[llm]
provider = "ollama"
model = "Qwen2.5:3b"
api_base = "https://ollama.example.com"
max_tokens = 1024
`

$3

`bash
OPENAI_API_KEY=sk-...
ANTHROPIC_API_KEY=sk-ant-...
`

Usage

`bash


From stdin (typical usage)

cat session.txt | llm-summarize --stdin

From clipboard

pbpaste | llm-summarize --stdin

Direct text

llm-summarize "session transcript text"
`

Options

| Flag | Description |
|------|-------------|
| --model | Override model from config |
| --max-tokens | Max output tokens (default: 1024) |
| --stdin | Read text from stdin |
| -h, --help | Show help |

Output

`json
{
"insights": {
"summary": "Implemented Redis caching layer with TTL and tag-based invalidation.",
"decisions": [
"Chose Redis over in-memory caching for persistence across restarts"
],
"patterns_used": [
"Tag-based cache invalidation"
],
"problems_solved": [
"Added caching to reduce database load with automatic invalidation on writes"
],
"tools_heavy": [
"Redis",
"CacheService"
]
},
"model": "Qwen2.5:3b",
"tokens_used": 126
}
`

$3

| Field | Description |
|-------|-------------|
| summary | One sentence: what was accomplished (always present) |
| decisions | Specific decisions with reasoning |
| patterns_used | Development patterns observed |
| preferences_expressed | User preferences revealed |
| problems_solved | Problems addressed and how |
| tools_heavy | Tools used notably |

Fields are omitted when no clear evidence exists in the transcript.

Supported Providers

| Provider | Models | API Key |
|----------|--------|---------|
| ollama | Qwen2.5:3b, llama3.2:3b, etc. | Not needed |
| anthropic | claude-3-5-haiku-latest, claude-sonnet-4-20250514 | Required |
| openai | gpt-4o-mini, gpt-4o | Required |

$3

`toml
[llm]
provider = "ollama"
model = "Qwen2.5:3b"
api_base = "https://ollama.example.com"
max_tokens = 1024
`

$3

`toml
[llm]
provider = "anthropic"
model = "claude-3-5-haiku-latest"
api_key = "env:ANTHROPIC_API_KEY"
max_tokens = 1024
`

Library Usage

`typescript
import { summarize, loadConfig, type SessionInsights } from "@voidwire/llm-summarize";

const config = loadConfig();
const result = await summarize("session transcript", config);

if (result.insights) {
console.log(result.insights.summary);
console.log(result.insights.decisions);
}
``

Exit Codes

| Code | Meaning |
|------|---------|
| 0 | Success |
| 1 | API error (rate limit, auth, network, parse failure) |
| 2 | Client error (missing args, invalid config) |