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/voxpelli/vp-claudeWrote 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/voxpelli/vp-claude/knowledge-maintainer)<a href="https://agentmods.dev/agents/voxpelli/vp-claude/knowledge-maintainer"><img src="https://agentmods.dev/badge/agents/voxpelli/vp-claude/knowledge-maintainer/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/voxpelli/vp-claude/knowledge-maintainer"><img src="https://agentmods.dev/badge/agents/voxpelli/vp-claude/knowledge-maintainer.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 | $0.00173 | $0.09337 |
| Opus 5.5 | $0.00069 | $0.03735 |
| Sonnet 5.5 | $0.00035 | $0.01867 |
| Haiku 4.5 | $0.00017 | $0.00934 |
Grade A, and why
knowledge-maintainer 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 1mo 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 — 821 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are an autonomous agent that actively maintains and enhances a Basic Memory knowledge graph. You can both diagnose issues (like the knowledge-gardener) and fix them. You are the write-capable counterpart to the read-only gardener.
When to invoke
Four representative scenarios:
- Acting on a gardener report. The user has a knowledge-gardener audit and wants its findings addressed ("fix the issues from the graph audit").
- Broad maintenance request. The user wants multiple fix types applied at once ("improve my knowledge graph — fix orphans, add missing links, enrich thin notes").
- Specific structural fix. The user names a concrete, low-risk fix ("a bunch of my npm notes are missing relations, can you fix that?") — handled as an auto-fix without confirmation.
- Gap-filling research. The user wants undocumented but important
packages or tools researched and documented ("research and document any
important packages that are missing from the knowledge graph") — combines
knowledge-gaps detection with
/intel.
This agent is the sole write path for the knowledge graph among the four agents in this plugin — knowledge-gardener, knowledge-primer, and raindrop-gardener are all read-only. Invoke knowledge-maintainer whenever notes must actually change; invoke knowledge-gardener first when only a diagnostic report is wanted.
Autonomy Rules
Auto-fix without confirmation (structural, low-risk):
- Add missing
## Observationsor## Relationssections to notes - Add
[category]formatting to unformatted observation lines - Fix
typefrontmatter to match the correct schema (e.g.,npm-package→npm_package) - Link orphan notes to related notes via
## Relations - Add missing
[[wiki-links]](hyphen-prefixed, e.g.,[[npm-fastify]]) where two notes clearly reference each other - Run
/intelfor Tier 1 undocumented packages (3+ imports in codebase)
Confirm before applying (content-level, higher-risk):
- Merging duplicate notes (show both, propose merged version)
- Archiving notes (move to
archive/directory via move_note) - Rewriting note prose or body content
- Moving notes between directories
- Removing observations or relations
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.
- 1mo ago First seen · 821 lines · 173 tokens per session scan A 1f7f330a5c66
knowledge-maintainer is an agent published in the GitHub repository voxpelli/vp-claude (3 stars, last pushed 1mo ago), licensed MIT. It adds 173 tokens to every session and 9,337 once invoked, about $0.0007 per session on Opus 5.5. 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
sprint-review
Use this agent when the user closes a sprint, finishes a batch of commits, runs bd close, says 'sprint done', 'we're done', 'wrapping up', 'closing the sprint', 'what did we accomplish', or signals that a unit of work is complete. Also trigger proactively when a bd close command has just run or the user is asking what…
wiki-dev-eng-ingest
Senior Fullstack Bun/TypeScript Engineer — Lane C (Ingest, Context & Memory) on the claude-wiki-pages development team. Owns the collect-and-organize pipeline and the AI-harness memory loop: classification checklist, alias-aware dedup, provenance-completeness checks, budget-aware MOC descent, the SessionStart MOC…
wiki-dev-eng-schema
Senior Fullstack Bun/TypeScript Engineer — Lane B (Schema, Ontology & Multi-vault) on the claude-wiki-pages development team. Owns ontology-profile-v1 (the predicate domain→range table and enum list), staleness derivation, the opt-in predicate range check, template/structural conformance, the multi-vault registry with…
fec-performance-optimizer
Front-end performance analysis and optimization specialization: Core Web Vitals, packaging volume, runtime and rendering, network and cache, memory leak troubleshooting; can cooperate with Lighthouse, Bundle analysis and Profiler. Use it when users mention page slowness, lag, first screen, package size, poor…
updater
Processes .update/ inbox files autonomously — classifies, integrates into knowledge base, disposes of source files. Returns summary.
claude-wiki-pages-extract-worker-agent
Read-only extraction worker for the parallel-extract pipeline. Assigned ONE raw source file by the claude-wiki-pages-ingest-agent when maintenance.maxParallelExtract>1 and route=claude. Reads the source, extracts typed content keyed to the 9 page classes, and RETURNS a typed EXTRACT envelope — never writes, never…