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 skills add LogicLabsAI/ultramemory-mcp --skill harnessgit clone --depth 1 https://github.com/LogicLabsAI/ultramemory-mcpWrote 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/skills/logiclabsai/ultramemory-mcp/harness)<a href="https://agentmods.dev/skills/logiclabsai/ultramemory-mcp/harness"><img src="https://agentmods.dev/badge/skills/logiclabsai/ultramemory-mcp/harness/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/logiclabsai/ultramemory-mcp/harness"><img src="https://agentmods.dev/badge/skills/logiclabsai/ultramemory-mcp/harness.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.00068 | $0.00386 |
| Opus 5 | $0.00034 | $0.00193 |
| Sonnet 5 | $0.00014 | $0.00077 |
| Haiku 4.5 | $0.00007 | $0.00039 |
Grade A, and why
harness 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 12d 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
Harness — one command, full rails
Task: $ARGUMENTS
Run these stages IN ORDER. Never start building before Stage 2 approval.
1 — Ground the plan (no memory)
Follow atomic-level-checklist-md-confirm-approval-needed for the task above: verify every fact
(paths, literal strings, schemas, endpoints) against actual source — never from memory — and emit a
machine-readable checklist in checklist-bound-execution format: each item =
{id, title, files[], spec, acceptance_criteria[], status}.
2 — STOP for approval (hard gate)
Present the checklist AND the proposed global gate command (e.g. npm run build && npm test).
Do NOT proceed until the user explicitly approves both the plan and the gate command.
3 — Execute through the harness
Follow checklist-bound-execution. Its step 0 arms the deterministic gate by writing the approved
command to .claude/.harness-active, then it embeds the checklist verbatim in the Workflow template,
partitions by file, and runs worker -> verifier with bounded loop-back using agentType
checklist-worker / checklist-verifier.
4 — Report
Summarize each item's final verdict plus any deviations or blocks. On turn end, the global Stop
hook runs the gate: on failure it blocks the stop and feeds back exactly what to fix; on green it
disarms automatically. You are not done until the gate is green.
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.
- 12d ago First seen · 36 lines · 68 tokens per session scan A c7923b0353e5
harness is a skill published in the GitHub repository LogicLabsAI/ultramemory-mcp (1 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 68 tokens to every session and 386 once invoked, about $0.0003 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
dashboard
Open OwnMem Console, the local dashboard for this repository's memory. Use when the user asks to open the dashboard, see memory metrics, check adoption or recall quality, or set up the optional embedding lane. Requires a repository initialized with the dashboard layer.
recall
Recall this repository's OwnMem local memory before changing code, and keep it healthy. Use when a repository contains .ownmem/, when past debugging lessons could apply ("have we hit this before", "why is it done this way"), or when the user mentions ownmem, project memory, or recalling across sessions.
init
Install or update OwnMem in the current repository. Use when the user asks to set up OwnMem, add local project memory for coding agents, or refresh an existing OwnMem installation after a version bump.
memory-write
Captures a durable fact to project memory through the kit's safe write path (PoisonGuard secret/injection screening + home-path sanitization + dedup + conflict detection) — preferring the cmk MCP tools (mkremember / mkforget / mktrust) when connected, falling back to the cmk CLI. Use when the user says "remember…
bootstrap
Scaffolds the core-memory-kit directory structure (context/, scripts/, cron/jobs/) into the current project. Idempotent — never overwrites existing files. Run once per project after installing the plugin. Use when the user says "bootstrap the memory system", "set up memory here", or "scaffold the memory kit".
troubleshooting
Diagnoses and repairs core-memory-kit itself when the memory system is broken — the per-failure-code repair book. Fire whenever a "⚠ [core-memory-kit]" whisper appears on the prompt (it names the failure and points here), or when the user reports a kit symptom however they phrase it: memory is not being saved, nothing…