Getting it into your agent
This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.
/plugin marketplace add andisab/swe-marketplace/plugin install knowledgewareWrote 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/commands/andisab/swe-marketplace/kb-maintain)<a href="https://agentmods.dev/commands/andisab/swe-marketplace/kb-maintain"><img src="https://agentmods.dev/badge/commands/andisab/swe-marketplace/kb-maintain/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/commands/andisab/swe-marketplace/kb-maintain"><img src="https://agentmods.dev/badge/commands/andisab/swe-marketplace/kb-maintain.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 | $0.00041 | $0.00416 |
| Opus 5.5 | $0.00016 | $0.00166 |
| Sonnet 5.5 | $0.00008 | $0.00083 |
| Haiku 4.5 | $0.00004 | $0.00042 |
Grade A, and why
kb-maintain 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 today.
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
Run a maintenance sweep against the knowledgebase site at: $ARGUMENTS
Read ${CLAUDE_PLUGIN_ROOT}/agents/kb-maintainer.md and execute its runbook in this session (not as a subagent — you need the Agent tool to fan out research subagents, which run on Sonnet). The governing contract is ${CLAUDE_PLUGIN_ROOT}/skills/knowledgebase/references/maintenance.md; its §5 protocol is binding.
Summary of what you will do (the runbook is authoritative):
- Parse the site's maintenance page — the master volatility table is your work order.
- Compute the due-list by cadence. Touch nothing that is not due.
- Fan out research subagents on Sonnet, clustered by authoritative source.
- Apply confirmed deltas surgically: research file → fan-out pages → earned stamps → one changelog row. With
--dry-run, report instead of writing. - Report drift as suggestions only — never restructure or rewrite the site.
If $ARGUMENTS is empty, ask for the site directory rather than guessing.
Scheduling recipe
Weekly headless run (cron, launchd, or any scheduler):
claude -p "/knowledgeware:kb-maintain '/path/to/site'" \
--permission-mode acceptEdits --model opus
Add --dry-run inside the quoted command for a report-only pass. Point one job per site; the sweep is self-scoping, so running it more often than weekly is safe — non-due rows are skipped, and a no-delta sweep still writes its changelog row.
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.
- today First seen · 30 lines · 41 tokens per session scan A 630af5e6a3e7
kb-maintain is a command published in the GitHub repository andisab/swe-marketplace (21 stars, last pushed today), licensed MIT. It adds 41 tokens to every session and 416 once invoked, about $0.0002 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 2026-10-17.
Other commands, from other repositories
ccc-update-docs
Regenerate the generated parts of project docs (scripts reference, environment variables, contributing guide, runbook) from source-of-truth files and flag stale docs, leaving hand-written prose intact. Use when docs drift from the code or after changing scripts, env vars or deploy steps.
report
Generate a publication-grade HTML report from a markdown source using the TRAID Design System catalog (5 templates). Agy matches content to template, you confirm, agy produces final branded HTML.
review-skills
Review changed skills: automated bash checks + 9 structural dimensions (D1-D9) + 5 intent checks (M1-M5). PASS/FAIL verdict. Fix in-place.
ccc-peers
Discover other running Claude Code instances and message them through the Claude Peers MCP server: list, send, broadcast, status, summary and inbox check. Use when several Claude Code sessions on one machine need to coordinate.
ccc-xray
Score the current project 0-100 across security, testing, DevOps, quality, docs and architecture, assign a maturity level, and rank skill recommendations by impact. Use when you want a quick project health scan or a list of what to fix first.
ccc-compound
Capture the patterns, corrections and decisions from the work just done into tasks/lessons.md, or review and prune existing lessons. Use after finishing a task or when the user says compound, lessons or retro.