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 Knuckles-Team/epistemic-graph --skill kg-modality-consensusgit clone --depth 1 https://github.com/Knuckles-Team/epistemic-graphWrote 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/knuckles-team/epistemic-graph/kg-modality-consensus)<a href="https://agentmods.dev/skills/knuckles-team/epistemic-graph/kg-modality-consensus"><img src="https://agentmods.dev/badge/skills/knuckles-team/epistemic-graph/kg-modality-consensus/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/knuckles-team/epistemic-graph/kg-modality-consensus"><img src="https://agentmods.dev/badge/skills/knuckles-team/epistemic-graph/kg-modality-consensus.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00133 | $0.00978 |
| Opus 5 | $0.00067 | $0.00489 |
| Sonnet 5 | $0.00027 | $0.00196 |
| Haiku 4.5 | $0.00013 | $0.00098 |
Grade A, and why
kg-modality-consensus 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 — 70 lines — stays where its author put it; the contents beside it link to each section on GitHub.
kg-modality-consensus — distributed consensus, resharding & tenants
The engine's durability/distribution layer makes the AI-native database horizontally
scalable and crash-safe. It runs openraft replication with automatic failover
(raft/cluster features, src/raft/), multi-Raft groups (an N-group ring with a
GroupRouter, KG-2.266/267/268), online resharding (reshard.rs moves ownership live,
hibernate/rehydrate groups), and cross-shard 2PC (presumed-abort, crash-recoverable) with
parallel-commit / read-only-participant fast paths and a Paxos-Commit-lite (and opt-in Calvin
deterministic) commit branch. A cross-region async read-replica tier with per-tenant quota +
backpressure guardrails rounds it out. See docs/capabilities.md → Durability & distribution.
The MCP way (through graph-os)
load_tools(tools=["engine_consensus", "engine_resharding", "engine_tenants",
"engine_rbac", "engine_admin"])
# engine_consensus — Raft membership / leadership / cluster health / failover
# engine_resharding — multi-Raft group ring, online ownership move, hibernate/rehydrate
# engine_tenants — per-tenant catalog: provision, list, route
# engine_rbac — RBAC policy administration: roles + resource/action grants
# engine_admin — ops/maintenance: online backup + restore
or the REST twins graph-os exposes for these engine domains.
Admin tier (scope-gated)
All five domains above are classified ADMIN (agent_utilities.mcp.tools.engine_tools. ADMIN_DOMAINS): every action they expose is denied fail-closed to an acting identity
that lacks the kg:admin scope/role (_enforce_action_scope), never merely hidden.
engine_rbac/engine_admin complete that ADMIN set — RBAC policy administration
(roles + resource/action grants) and the ops backup/restore surface — so use this
skill whenever you need to grant/revoke a role, inspect a resource's ACL, or run an
online backup/restore, exactly like reaching for engine_consensus/engine_tenants
for cluster/tenant admin.
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 · 70 lines · 133 tokens per session scan A 00b487ba91c1
kg-modality-consensus is a skill published in the GitHub repository Knuckles-Team/epistemic-graph (10 stars, last pushed 12d ago), licensed MIT. It adds 133 tokens to every session and 978 once invoked, about $0.0007 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 skills, from other repositories
cortivex-consensus
Raft-style leader election for multi-node Cortivex clusters using ConsensusManager nodes.
data-search-engine
Use this skill when asked about Elasticsearch, OpenSearch, Solr, search engine, full-text search, inverted index, indexing, search analytics, aggregation, cluster management, or shard routing. This skill enforces: Elasticsearch architecture (node types, shards, replicas), indexing strategy (mapping, analysis…
documentdb-collection-admin
Database, collection, and user administration on a DocumentDB (MongoDB-compatible, PostgreSQL-backed) server via the documentdb-mcp MCP server — list/create/drop/rename databases and collections, and manage database users and their roles. Use when the agent must provision or tear down namespaces, enumerate the…
django
Operational skill for Django: models, ORM query hygiene, migrations, views/URLs, settings security, and admin customization.
analytics-data-manager
Use this skill when configuring Data Manager in CRM Analytics: enabling objects for sync, scheduling data sync runs, setting up remote connections to external databases, monitoring sync status and error logs, or troubleshooting connected object issues. Trigger keywords: data sync, connected objects, Data Manager, sync…
compound-field-patterns
Compound fields (Name, Address, Geolocation): SOQL access rules, DML semantics, component access in Apex/LWC, reporting column behavior, formula field restrictions. NOT for creating a new custom field — use admin/custom-field-creation. NOT for formula syntax and functions — use admin/formula-fields.