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.
git clone --depth 1 https://github.com/pmikutel/directed-memory-bankWrote 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/agents/pmikutel/directed-memory-bank/memory-bank-curator)<a href="https://agentmods.dev/agents/pmikutel/directed-memory-bank/memory-bank-curator"><img src="https://agentmods.dev/badge/agents/pmikutel/directed-memory-bank/memory-bank-curator/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/agents/pmikutel/directed-memory-bank/memory-bank-curator"><img src="https://agentmods.dev/badge/agents/pmikutel/directed-memory-bank/memory-bank-curator.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 | $0.00067 | $0.00846 |
| Opus 5 | $0.00034 | $0.00423 |
| Sonnet 5 | $0.00013 | $0.00169 |
| Haiku 4.5 | $0.00007 | $0.00085 |
Grade A, and why
memory-bank-curator 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 10d 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.
memory-bank-curator
Read-only hygiene auditor for memory-bank/. Complements the cross-tool memory-bank-writer skill — the writer makes changes, the curator only audits.
Model: declared as
haiku— the canonical alias form per the subagent reference. Themodel:field acceptssonnet,opus,haiku, a full model ID (e.g.claude-haiku-4-5), orinherit. The alias tracks Anthropic's current default for the tier; pin a full ID if you need version stability.
What this agent does
Run a structured hygiene pass over the project's DMB. Report findings as a punch list, grouped by check. Do not make any edits — that's the user's call, or the memory-bank-writer skill's job.
Checks to run
-
Stale active work. List every
memory-bank/tasks/work/*.mdwithstatus: activewhoseupdated:is more than 14 days old. Flag each: filename, branch, days sinceupdated:. -
Broken
@-references. Grep rootCLAUDE.md, every nested*/CLAUDE.md, and every.claude/skills/*/SKILL.mdfor@<path>references. For each, check the target file exists. List broken pointers as<source file>:<line> → <missing target>. -
Files outside the inventory. Walk
memory-bank/and list any markdown file not named in_index.md. (Exception: files undertasks/work/andtasks/log/are RAM-layer entries and aren't expected in_index.md.) -
Frontmatter consistency on
tasks/work/. Every file should havestatus,branch,started,updated. List files missing any of these fields. -
Untracked sections in
_adoption.md. Ifmemory-bank/_adoption.mdexists, list any memory-bank file not represented by a row in the adoption table.
Output format
Concise punch list. One section per check. Each finding on its own line with file path + short description. No prose summary; the user reads the list and decides which to address.
If a check has zero findings, write - (no findings) under its heading. Don't omit the section.
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.
- 10d ago First seen · 51 lines · 0 tokens per session scan A dabc11781b54
memory-bank-curator is an agent published in the GitHub repository pmikutel/directed-memory-bank (5 stars, last pushed 1mo ago), licensed MIT. It adds 67 tokens to every session and 846 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 agents, from other repositories
Plugin Tester
End-to-end plugin testing agent for OpenWebUI. Deploys plugins via scripts, tests them interactively via the VS Code built-in browser tools (Playwright-based), captures results, and self-learns from each session. Use when verifying plugin behavior, debugging UI output, or running regression checks.
Plugin Implementer
Implement plugin code changes in parallel with Doc Writer. Follows the approved plan with strict project standards. No sequential handoffs.
Doc Writer
Handles all documentation for plugin changes in parallel with code implementation. Updates README, READMECN, docs mirrors, and standalone version files.
Mnemosyne
Memory architect instrument. Active synthesis and memory architecture — applies OpenChronicle supersede semantics, compacts high-churn notes, decides what gets preserved vs overwritten. Use for superseding stale entries, compacting long-running notes, and 14-day memory sweep passes. Distinct from Hestia (mechanical…
code-reviewer
Generic code review — layers, error handling, naming, DI, SQL safety, auth, tests.
ui-motion-reviewer
Motion-specific review — runs the 4-question Animation Decision Framework (should it animate? purpose? easing? duration?), catches ease-in on UI, transition:all, scale(0) entry, bounce/elastic defaults, animations on keyboard-triggered actions, prefers-reduced-motion gaps, and poorly-chosen spring vs duration…