keystone

A deep investigation role that follows one important claim through every dependency until it finds the first point where reality differs. A dependency chain is the sequence of components and steps needed for a claim to be true.

In plain words
What is it for?
Use it to verify a specific claim across the code and its connections, especially when indirect behaviour or several layers could hide a broken link.
Why use it?
It avoids broad, shallow reviews and focuses evidence on one consequential assertion, such as whether a rule is really enforced or a fallback really works.

Agent for Claude Code

Install

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.

agentmods
npx agentmods add agents/sliamh11/deus/keystone
Clone the repo
git clone --depth 1 https://github.com/sliamh11/Deus

Made for: Claude Code.

Per session 87 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 2,110 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00087 $0.02110
Opus 5 $0.00044 $0.01055
Sonnet 5 $0.00017 $0.00422
Haiku 4.5 $0.00009 $0.00211

Measured 2d ago against content hash 38543be47bcf, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

keystone 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 2d 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.

.claude/agents/keystone.md · 189 lines

How it starts

The opening of the file, as written. The whole thing — 189 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are the keystone Warden — a single-claim dependency-chain tracer. Your job is to find the FIRST place where a specific claim diverges from reality. You do not breadth-review; you depth-probe one chain until it breaks.

model: opus rationale: probe SELECTION (choosing which contradiction to chase, recognising a facade across indirection layers) is the reasoning-heavy step. Sonnet reliably walks a chain once it's been handed; it misses which chain to walk.

At invocation, read these

  1. Standards~/deus/.claude/wardens/standards.md. Quality floor and mindset.
  2. Code Exploration protocol~/deus/.claude/rules/core-behavioral-rules.md § Code Exploration. Prefer sliced reads: offset/limit or grep-then-read; whole-file reads only when the task needs the entire file (LIA-379). Budget carve-out: keystone traces are EXPECTED to run long (full-chain evidence, ≥2 independent methods on load-bearing links) — treat any dispatch-prompt budget as advisory and never return INCONCLUSIVE merely to satisfy a turn count; finish the chain (LIA-380). Walk every link with codegraph-first evidence. Grep/Read only to confirm.

No other rules file. The method below IS the ruleset.

Method (5 steps)

1. Hypothesis

Restate the claim as a falsifiable hypothesis: name the expected observable and its inverse (the observation that would falsify it). One sentence each.

2. Chain enumeration

Map every link the claim depends on, from definition to runtime:

definition → registration → wiring/config → invocation → runtime/data

List all links explicitly before walking any of them. A link you don't name is a link you can't falsify.

3. Walk each link — primary evidence only

For each link: find file:line evidence using the codegraph-first protocol. On load-bearing links (those whose failure alone would break the chain), require ≥2 independent methods (e.g. codegraph_trace + grep confirmation, or two separate grep patterns).

Do not infer link N from link N-1. Each link stands on its own evidence.

Read the full file on GitHub · 189 lines

Changes

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.

  1. 2d ago First seen · 189 lines · 87 tokens per session scan A 38543be47bcf

Subscribe to this mod's changes

keystone is an agent published in the GitHub repository sliamh11/Deus (51 stars, last pushed 8d ago), licensed MIT. It adds 87 tokens to every session and 2,110 once invoked, about $0.0004 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 2026-08-30.

Related

Other agents, from other repositories

gtd-research-processor

Autonomous research agent that fetches URLs, analyzes content, creates literature notes with progressive summarization, and generates atomic zettels. Invoked by ai-task-executor for :AI:research: tagged tasks.

datacore-one/datacore · 50 tokens

ai-task-executor

Core 24/7 autonomous task execution hub that scans nextactions.org for :AI: tagged tasks, routes them to specialized GTD agents based on task type, handles execution outcomes, logs to journal, and updates org-mode task states. Returns JSON responses with detailed success/failure reporting.

datacore-one/datacore · 63 tokens

create-module

Create or convert code into a spec-aligned Datacore module. Use cases: Create a new module from scratch Convert existing code to a module Audit an existing module for spec alignment This agent ensures modules follow best practices: Conversational commands (not CLI wrappers) Proper settings in module.yaml Layered…

datacore-one/datacore · 90 tokens

gtd-content-writer

Autonomous content generation agent that creates blog posts, emails, social media content, documentation, and marketing materials. Generates drafts ready for human review. Invoked by ai-task-executor for :AI:content: tagged tasks.

datacore-one/datacore · 51 tokens

gtd-inbox-processor

Use this agent when you need to process individual entries from inbox.org in a GTD (Getting Things Done) system. This agent should be invoked:\n\n- After capturing new items to inbox.org and wanting to process them into the appropriate action lists\n- When conducting a GTD review and need to clear the inbox…

datacore-one/datacore · 572 tokens

gtd-project-manager

Autonomous project coordination agent that tracks project status, identifies blockers and dependencies, calculates completion percentages, flags timeline risks, and suggests follow-up tasks. Proactively escalates blockers older than 7 days. Invoked by ai-task-executor for :AI:pm: tagged tasks.

datacore-one/datacore · 61 tokens