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 brodewa369/roowet-semantic-vault-search --skill claim-update-gategit clone --depth 1 https://github.com/brodewa369/roowet-semantic-vault-searchWrote 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/brodewa369/roowet-semantic-vault-search/claim-update-gate)<a href="https://agentmods.dev/skills/brodewa369/roowet-semantic-vault-search/claim-update-gate"><img src="https://agentmods.dev/badge/skills/brodewa369/roowet-semantic-vault-search/claim-update-gate/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/brodewa369/roowet-semantic-vault-search/claim-update-gate"><img src="https://agentmods.dev/badge/skills/brodewa369/roowet-semantic-vault-search/claim-update-gate.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 | $0.00018 | $0.00549 |
| Opus 5.5 | $0.00007 | $0.00220 |
| Sonnet 5.5 | $0.00004 | $0.00110 |
| Haiku 4.5 | $0.00002 | $0.00055 |
Grade A, and why
claim-update-gate 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
Claim Update Gate — no dualism in the vault
When to Use
Before writing ANY stateful claim about a project, tool, or entity to the vault — project state, config state, tool status, entity status. Not for incident logs (those stay dated).
Why
Benchmark 2026-10-17 (cron topic) proved lexical RAG surfaces stale claims as present-tense truth. Root cause is write-side: the agent creates NEW files for facts that already have an owner file. Two files arguing about the same present-tense fact = dualism noise that ranking cannot fix.
The gate (run BEFORE every vault write)
- Classify the claim:
- Stateful — project status, tool/config state, entity state ("stokbot is on Supabase pooler", "cron model is mimo @9rt").
- Incident/dated — error, lesson, decision, daily note ("hit error X on 2026-10-17").
- STATEFUL claims:
search_vault("<entity/topic>")first.- Owner file exists (project overview, LOG.md, tool note, entity note)? → UPDATE that file: append the new claim to its
## Historysection with date; never delete prior entries. - No owner file → create ONE state file for that entity; it becomes the single owner from now on.
- Creating a second state file for a fact that has an owner is a violation.
- INCIDENT/dated claims:
- New date-stamped file (existing per-type rules).
- MUST backlink the stateful owner file it changes status of (## Related), AND update that owner file's History with a one-line pointer to the incident file.
- Never merge multiple entries into one file (existing rule). History section = dated one-liners, not prose blobs.
Enforcement
- Self-check before reporting done: did this task create any state file whose fact already had an owner? If yes → merge into owner; delete nothing without confirmation.
- Decision record: [[01-AGENT-MEMORY/decisions/2026-10-17-claim-update-protocol-write-gate]]
Related
- Skills: vault-management, rag-vault-search
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 · 41 lines · 18 tokens per session scan A b5418d114e64
claim-update-gate is a skill published in the GitHub repository brodewa369/roowet-semantic-vault-search (5 stars, last pushed today), licensed MIT. It adds 18 tokens to every session and 549 once invoked, about $0.0001 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 skills, from other repositories
recallnest
Recover cross-session context with RecallNest, search durable memory before repeat exploration, store durable decisions and lessons, and checkpoint unfinished work.
swarmvault
Use SwarmVault when the user needs a local-first knowledge vault that writes durable markdown, graph, search, dashboard, review, chat-session, context-pack, task-ledger, static AI export, retrieval, and MCP artifacts to disk from books, notes, transcripts, exports, datasets, slide decks, files, URLs, code, and…
recallnest
Recover cross-session context with RecallNest, search durable memory before repeat exploration, store durable decisions and lessons, and checkpoint unfinished work.
wiki-save
A tool for saving useful parts of the current conversation as organised wiki pages. It classifies the material, adds page metadata and links, and updates the wiki’s indexes and activity records.
domain-status
Summarize the current state of one vault domain (CloudOps, DevOps, FinOps, SecOps, Architecture) - its notes, what changed recently, and the open loops that name it. Use before a weekly review, before picking up work in a domain, or when asked "where are we on X".
session-context
Load the vault's Index/ files (Vault Map, Active Context, Open Loops) so the session starts from the real state instead of re-deriving it. Use at the start of a session, after a context switch, or when an agent's answers drift from what the vault says.