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 skills/u4csolutions/midmem/midmem-recordnpx skills add U4CSolutions/MidMem --skill midmem-recordgit clone --depth 1 https://github.com/U4CSolutions/MidMemWhat 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 | $0.00105 | $0.00846 |
| Opus 5 | $0.00053 | $0.00423 |
| Sonnet 5 | $0.00021 | $0.00169 |
| Haiku 4.5 | $0.00011 | $0.00085 |
Grade A, and why
midmem-record 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 — 51 lines — stays where its author put it; the contents beside it link to each section on GitHub.
MidMem Record — durable lessons + commit, done right
The point: durable knowledge must land in state.db, not just the chat transcript, so it survives
context compaction and is recallable by every agent on the shared store.
Steps
- Lesson → midmem (the load-bearing step). Store each lesson atomically:
midmem remember "<the lesson, one sentence>" --tier wisdom --curated --scope shared- ⚠️ Escaping gotcha: the text must not contain backticks,
$(...), or unescaped<> |— the shell runs them as substitution and mangles/blanks the entry. Plain prose, no backticks; replace code refs with words ("call query with a token budget", not the backticked form). One per lesson.
- ⚠️ Escaping gotcha: the text must not contain backticks,
- Changelog (if your project keeps one). Add under today's dated
## YYYY-MM-DDsection (newest at top), grouped: Added · Changed · Fixed · Reverted · Removed · Failed Attempts · Lessons (omit empty groups). Be specific with values/paths ("rrfK 60→45", not "tuned retrieval").- ⚠️ One dated section per day, one of each group per day — append to today's existing groups;
don't create a second
## <date>or a second### Lessons.grep -nE '^## |^### ' CHANGELOG.mdshould show no duplicate date/group headers for one day.
- ⚠️ One dated section per day, one of each group per day — append to today's existing groups;
don't create a second
- Vault projection.
midmem projectrefreshes the LLM Wiki (or let lifecycle maintenance auto-project); only force if you need it visible immediately. Never hand-edit the projection — it is regenerated fromstate.db. - Commit. Stage the changed files (+ changelog if touched). End the message with the active
model's
Co-Authored-By:line. Verify the staged diff carries no secrets —state.db,*.pem,.env, and service-account JSON should be gitignored; check before committing.
Harness-guaranteed recording (the reliable pattern)
Instruction-following alone loses records — a session ends, the lesson evaporates. Make it
deterministic: wire a Claude Code Stop hook (via .claude/settings.json) that blocks a turn
from ending while a recordable change is unrecorded, clearing when midmem remember / midmem ingest runs. A companion PostToolUse hook flags the session "dirty" when a mutating command runs.
This turns "please remember to record" into a gate the session cannot skip. Make it fail open on
error and yield after a few blocks so it never wedges a session. (This is the model behind an
operator's host record skill; this repo copy keeps the portable pattern without the host-specific
changelog/vault plumbing.)
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 · 51 lines · 105 tokens per session scan A b04f069938e7
midmem-record is a skill published in the GitHub repository U4CSolutions/MidMem (2 stars, last pushed 8d ago), licensed Apache-2.0. It adds 105 tokens to every session and 846 once invoked, about $0.0005 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 skills, from other repositories
potpie-source-ingestion
Use when the user explicitly asks to ingest, refresh, or deeply understand a repository, PR, issue, ticket, runbook, incident report, document, or web link into Potpie. The harness performs todo-driven discovery, uses local/GitHub/integration tools and read-only subagents when available, builds evidence-backed…
potpie-repo-baseline
Use when establishing, refreshing, or deeply understanding a repository's baseline memory in Potpie: purpose, application type, features, services/modules, environments, deploy shape, dependencies, API contracts, datastores, integrations, ownership, and explicit preferences. The harness reads authored and…
graph-mutation-plan
Cookbook for composing an applygraphmutations plan — stable entitykey patterns, the canonical label/edge vocabulary, evidence/invalidation/confidence discipline, and a worked example. Load this when building a non-trivial mutation plan.
obsidian-layout-adjustment
Workflow for working with the user on changing how Obsidian looks using CSS snippets. Use this whenever the user asks to restyle Obsidian, tune a vault's visual layout, adjust tabs, sidebars, note surfaces, properties, backlinks, graph panes, file explorer rows, icons, links, shadows, active states, or CSS snippets.…
alfworld-locate-target-object
Navigates to a suspected location and identifies a target object. Use when your goal requires finding a specific object (e.g., "potato", "plate") and its location is not immediately known. Moves to a relevant receptacle (like a fridge or cabinet), checks its contents, and outputs the object's location or confirms its…
scienceworld-growth-focuser
Use when you have planted a seed or need to track a plant's growth stage (sprouting, flowering, reproduction). Applies the 'focus on' action to a specific plant or biological entity to signal intent and monitor its development. Trigger after planting or when you need to observe life cycle progression in the…