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 instructions/atazifor/engineering-memlog/claude-mdgit clone --depth 1 https://github.com/atazifor/engineering-memlogWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/instructions/atazifor/engineering-memlog/claude-md)<a href="https://agentmods.dev/instructions/atazifor/engineering-memlog/claude-md"><img src="https://agentmods.dev/badge/instructions/atazifor/engineering-memlog/claude-md.svg" alt="Measured on agentmods" height="20"></a>What 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 | $0.00993 | $0.00993 |
| Opus 5 | $0.00496 | $0.00496 |
| Sonnet 5 | $0.00199 | $0.00199 |
| Haiku 4.5 | $0.00099 | $0.00099 |
Grade A, and why
engineering-memlog CLAUDE.md 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 4d 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.
engineering-memlog plugin — agent instructions
This file is auto-loaded by the engineering-memlog plugin and forms the agent's working contract with the log. Two halves: read before you work, write after.
How the log gets into your context
The plugin closes the read side without relying on you to remember:
- SessionStart hook runs once per session, sniffs the project signature (languages, frameworks, repo, service hint), and injects the top 6 most relevant prior lessons as a system reminder you see at session boot.
- UserPromptSubmit hook scans each prompt for symptom-shaped text (errors, failures, HTTP 4xx/5xx, "broken", "not working", framework names) and, when triggered, automatically searches the log for likely keywords and injects matches before you respond. Silent on benign prompts — no per-turn noise.
/recall <query>slash command for explicit deep-dive search when you suspect something but the hooks didn't catch it.
When the hook injects entries, treat them as prior knowledge, not
documentation. Each line is a full JSON memlog entry. If a current
symptom matches one of the entries' problem or cause, apply its
prevention rule rather than re-deriving from scratch. Reference the
entry's id or title in your reasoning so the human can trace where
the rule came from.
Search the log yourself when warranted
The hooks fire automatically, but you should also run the log yourself when:
- The injected entries didn't include something you suspect exists.
- You hit an unfamiliar error during the work (not just at prompt time).
- You're about to propose a fix for a non-obvious bug.
Run it as a Bash command:
~/engineering-memory/bin/memlog search "frozen-lockfile" --json --limit 10
Search for the concrete signal — keywords from the error message, the observed symptom, the framework or tool involved. No match is normal — just proceed.
Append a lesson after you work
Before concluding a meaningful task, ask:
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.
- 4d ago First seen · 99 lines · 993 tokens per session scan A 181074e953a7
engineering-memlog CLAUDE.md is an instructions file published in the GitHub repository atazifor/engineering-memlog (1 stars, last pushed 2mo ago), licensed MIT. It adds 993 tokens to every session, about $0.0050 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 instructions, from other repositories
THOR-memory-bible-for-AI AGENTS.md
AGENTS.md instructions for nworks3d/THOR-memory-bible-for-AI, covering agents.md - read this if you are the ai assistant, setting them up, the tools you have, how to write a note that actually comes back and after setup, this happens without you.
memsearch CLAUDE.md
Claude Code instructions for zilliztech/memsearch, covering claude.md, build & test commands, install in development mode, run a single test file and run a specific test.
plur CLAUDE.md
Claude Code instructions for plur-ai/plur, covering claude.md, what is plur, development, package dependency and version bumps.
honcho CLAUDE.md
Claude Code instructions for plastic-labs/honcho, covering claude.md, honcho overview, what is honcho?, core concepts and peer paradigm.
altk-evolve AGENTS.md
AGENTS.md instructions for AgentToolkit/altk-evolve, covering what is evolve?, key concepts, architecture flow, project directory tree (some files omitted for brevity) and first time setup.
zer0dex AGENTS.md
Instructions for hermes-labs-ai/zer0dex, covering agents.md, use it for, do not use it for, minimal commands and output shape.