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/macbyclp/memorry/hatirlagit clone --depth 1 https://github.com/macbyclp/memorryWhat 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 | $0.00011 | $0.00457 |
| Opus 5 | $0.00005 | $0.00229 |
| Sonnet 5 | $0.00002 | $0.00091 |
| Haiku 4.5 | $0.00001 | $0.00046 |
Grade A, and why
hatirla 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.
What it actually says
The user wants this committed to persistent memory (memorry): $ARGUMENTS
Steps:
- Make sure the memorry server is running (port 8765). If it isn't, start
it — on Windows via
start-memory.bat, otherwisepython server.pyfrom your memorry checkout. - Distill the text into ONE self-contained sentence — one that still makes sense read on its own months from now, with no surrounding conversation. If the user already phrased it that way, keep their wording; don't pad it with commentary.
- Search the same topic with
memory_searchfirst. If you find a record that is now stale or contradicts the new fact, remove it withmemory_delete(or correct it in place withmemory_update) before writing. - Pick the right
project. Keep personal facts and preferences separate from infrastructure/setup decisions — see the note below on setting your own default. - Write a
titleof 3–6 words naming the topic (not a shortened copy of the sentence). It becomes the note's filename and its label in Obsidian's Graph View, so a truncated sentence there makes the graph unreadable. - Add 1–3
tags, reusing tags already in use — check withmemory_listif you don't know them. Tags are what group records into clusters and what the Graph View colours by; a new tag per record gives you neither. - Save it with
memory_add. - Reply with a single line: what you saved and the id it got. No essay.
Customize this: step 4's routing is deliberately generic. Edit this file to name your own projects (e.g.
personalvsinfra, or one per codebase) and to hardcode the path to yourstart-memory.batif you want step 1 to be a single copy-pasteable command.
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 · 38 lines · 11 tokens per session scan A c51598d60642
hatirla is a command published in the GitHub repository macbyclp/memorry (3 stars, last pushed 15d ago), licensed MIT. It adds 11 tokens to every session and 457 once invoked, about $0.0001 per session on Opus 5. 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 2026-08-31.
Other commands, from other repositories
git
Git operations with intelligent commit messages and workflow optimization.
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.