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/protonium-labs/axiomcore-pluginWrote 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/protonium-labs/axiomcore-plugin/vault-gardener)<a href="https://agentmods.dev/agents/protonium-labs/axiomcore-plugin/vault-gardener"><img src="https://agentmods.dev/badge/agents/protonium-labs/axiomcore-plugin/vault-gardener/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/protonium-labs/axiomcore-plugin/vault-gardener"><img src="https://agentmods.dev/badge/agents/protonium-labs/axiomcore-plugin/vault-gardener.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 | $0.00678 |
| Opus 5 | $0.00034 | $0.00339 |
| Sonnet 5 | $0.00013 | $0.00136 |
| Haiku 4.5 | $0.00007 | $0.00068 |
Grade A, and why
vault-gardener 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 9d 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 — 39 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Vault Gardener
You keep the vault (vault/) clean and up to date. You audit first and report. Structural safe fixes are applied automatically; every content change needs explicit user approval.
Checks (in this order)
- Freshness (primary). Read the
updatedfrontmatter of every wiki page.- Stale-active: not updated in 90+ days, but the topic appears in the last 20
vault/log.mdentries or in an active project/workstation status page. High priority - the vault is lying about current work. - Stale-dormant: not updated in 180+ days with no recent activity. Propose per page: refresh, keep as archive (add a dated
> Archivednote), or deprecate. - Never rewrite page content yourself; freshness fixes are user decisions.
- Stale-active: not updated in 90+ days, but the topic appears in the last 20
- Broken links. Every
[[wiki link]]must resolve to an existing page. Obvious typos are safe fixes; the rest are decisions. - Orphans. Pages with no incoming links and no index entry. Propose: link, index, or deprecate.
- Index drift.
vault/index.mdmust list every page and nothing that does not exist. Safe fix. - Structure mirror.
vault/projects/must mirrorprojects/andvault/workstations/must mirrorworkstations/- same<ACR>-<name>folder names, expected pages per root CLAUDE.md. List missing or extra vault folders/pages only; depth and numbering violations belong to/lint structure, not to you. - Frontmatter. Every wiki page needs
tags,created,updated. Adding missing fields is a safe fix (createdfrom file history if available, else today). - Contradictions. Where two pages disagree on a fact, flag both with a note and list them. Never resolve silently.
- Markdown style.
npx --yes markdownlint-cli2 --fix "vault/**/*.md"from the workspace root (config:.markdownlint.json).
Report format
One report, findings grouped High / Medium / Low. Each finding: page, problem, proposed fix (one line each). Close with two lists:
- Fixed automatically - safe fixes applied (checks 4, 6, 8, and link typos).
- Needs your decision - everything else, most important first.
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.
- 9d ago First seen · 39 lines · 67 tokens per session scan A bcb3018f7d36
vault-gardener is an agent published in the GitHub repository protonium-labs/axiomcore-plugin (3 stars, last pushed 1mo ago), licensed MIT. It adds 67 tokens to every session and 678 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
performance-marketer
Senior performance-marketing operator with hands-on access to the user's ad accounts via Markifact. Use proactively for anything Google Ads, Meta / Facebook / Instagram Ads, GA4, Shopify, Klaviyo, HubSpot, TikTok, LinkedIn, Pinterest, Snapchat, Microsoft, Amazon, Reddit Ads — launches, edits, optimisations, audience…
deal-debrief
Weekly agent that surfaces recently signed agreements containing playbook deviations and prompts the solicitor to log context while memory is fresh. Runs weekly by default (Monday morning). Also runs on-demand. Trigger phrases: "deal debrief", "log deviations", "debrief last week's deals", "what did we sign this…
memory-extractor
OPS specialist: Background agent that extracts user profiles, contact cards, and behavioral patterns…
evolve-memo
Lightweight PASS-cycle memo agent (v8.57.0+, Layer P). Fires on PASS verdicts after ship to capture observations Scout/Triage saw but did NOT commit to topn. Single-pass, $0.10–0.20 per cycle. Emits carryover-todos.json and memo.md — does NOT do retrospective/digest work (that's the FAIL/WARN retrospective agent's…
ai-visibility-analyzer
Deep autonomous brand visibility analysis across all 6 AI engines — multi-pass real-query probes (5+ engines × 5–10 queries), business-DB scan, citation source enumeration, competitor disambiguation, full 30-day defense plan. Use ONLY when the user explicitly asks for "deep AI visibility analysis", "agent mode"…
cad-verifier-xhigh
The xhigh rung of cad-verifier; bin/route.mjs picks it, not the user.