Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add commands/codenamev/claude_memory/distill-transcriptsgit clone --depth 1 https://github.com/codenamev/claude_memoryWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00000 | $0.00809 |
| Opus 5 | $0.00000 | $0.00404 |
| Sonnet 5 | $0.00000 | $0.00162 |
| Haiku 4.5 | $0.00000 | $0.00081 |
Grade A, and why
distill-transcripts scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 99 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Distill Transcripts
Extract structured knowledge (facts, entities, decisions) from undistilled transcript content and persist it to long-term memory.
Usage
/distill-transcripts
/distill-transcripts --limit 10
Instructions
You are a knowledge extraction specialist. Your job is to read raw transcript content and extract structured facts, entities, and decisions, then persist them via the memory.store_extraction MCP tool.
Step 1: Get Undistilled Content
Call memory.undistilled with limit: 10 to get transcript content that hasn't been processed yet.
If no items are returned, report "No undistilled content found" and stop.
Step 2: Extract Knowledge (per item)
For each content item, carefully read the raw_text and extract:
Entities — Named things mentioned:
- type: database, framework, language, platform, repo, module, person, service
- name: Canonical name (e.g., "PostgreSQL" not "postgres")
- confidence: 0.0-1.0
Facts — Knowledge learned:
- subject: Entity name or "repo" for project-level facts
- predicate: uses_database, uses_framework, convention, decision, auth_method, deployment_platform, depends_on, testing_strategy
- object: The value
- confidence: 0.0-1.0
- quote: Source excerpt (max 200 chars)
- strength: "stated" (explicitly said) or "inferred" (implied)
- scope_hint: "project" (this project only) or "global" (all projects)
Decisions — Choices made:
- title: Short summary (max 100 chars)
- summary: Full description
- status_hint: "accepted", "proposed", or "rejected"
What to Extract
- Technology choices ("we use PostgreSQL", "switched to React")
- Conventions ("always use frozen_string_literal", "test files go in spec/")
- Architectural decisions ("API uses REST", "auth via JWT")
- Preferences ("prefer 4-space indent", "use Standard Ruby")
- Project structure ("migrations in db/migrations/", "commands in commands/")
What to Skip
- Debugging steps and transient errors
- Code output and tool observations
- File contents that were just being read
- Ephemeral task details ("fix this test", "run the linter")
- Information already obvious from the codebase itself
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 99 lines · 0 tokens per session scan A 9fdaddeedc24
distill-transcripts is a command published in the GitHub repository codenamev/claude_memory (24 stars, last pushed 1mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 809 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other commands, from other repositories
checklist
Generate a custom checklist for the current feature based on user requirements.
clarify
Identify underspecified areas in the current feature spec by asking up to 5 highly targeted clarification questions and encoding answers back into the spec.
specify
Create or update the feature specification from a natural language feature description.
analyze
Perform a non-destructive cross-artifact consistency and quality analysis across spec.md, plan.md, and tasks.md after task generation.
converge
Assess the current codebase against the feature's spec, plan, and tasks, then append any remaining unbuilt work as new tasks to tasks.md so implement can complete it.
implement
Execute the implementation plan by processing and executing all tasks defined in tasks.md.