agents

Instructions for coding agents that can read documentation but cannot use slash commands. It describes the Clavix process, including checking applicable skills and following separate planning, implementation and verification stages.

In plain words
What is it for?
Use it to guide documentation-only agents through Clavix tasks, select the correct mode and follow required checks and handoffs.
Why use it?
It helps an agent choose the right workflow and avoid claiming work is complete without evidence. Some stages deliberately stop before code is written.

Agent

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/bobvibes/clavix/agents
Clone the repo
git clone --depth 1 https://github.com/bobvibes/Clavix
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 2,055 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 $0.00000 $0.02055
Opus 5 $0.00000 $0.01027
Sonnet 5 $0.00000 $0.00411
Haiku 4.5 $0.00000 $0.00205

Measured 3d ago against content hash 7ea23489d78e, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

agents 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 3d 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.

src/templates/agents/agents.md · 218 lines

How it starts

The opening of the file, as written. The whole thing — 218 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Clavix Instructions for Generic Agents

This guide is for agents that can only read documentation (no slash-command support). If your platform supports custom slash commands, use those instead.


🚨 FIRST: Load using-clavix Skill

Before ANY Clavix workflow, invoke the using-clavix skill.

This meta-skill establishes:

  • Skill invocation rules (check skills BEFORE any response)
  • Required skill chains (prd → plan → implement → verify)
  • Iron Laws for verification (no completion claims without evidence)
  • Workflow orchestration and fix loops

If you have even a 1% chance a Clavix skill applies, you MUST check it first.

In environments with Skill tool: Invoke using-clavix before any Clavix action. In environments without Skill tool: Read .config/agents/skills/using-clavix/SKILL.md or the bundled template.


⛔ CLAVIX MODE ENFORCEMENT

CRITICAL: Know which mode you're in and STOP at the right point.

OPTIMIZATION workflows (NO CODE ALLOWED):

  • Improve mode - Prompt optimization only (auto-selects depth)
  • Your role: Analyze, optimize, show improved prompt, STOP
  • ❌ DO NOT implement the prompt's requirements
  • ✅ After showing optimized prompt, tell user: "Run /clavix:implement --latest to implement"

PLANNING workflows (NO CODE ALLOWED):

  • Conversational mode, requirement extraction, PRD generation
  • Your role: Ask questions, create PRDs/prompts, extract requirements
  • ❌ DO NOT implement features during these workflows

IMPLEMENTATION workflows (CODE ALLOWED):

  • Only after user runs execute/implement commands
  • Your role: Write code, execute tasks, implement features
  • ✅ DO implement code during these workflows

If unsure, ASK: "Should I implement this now, or continue with planning?"

See .clavix/instructions/core/clavix-mode.md for complete mode documentation.


📁 Detailed Workflow Instructions

For complete step-by-step workflows, see .clavix/instructions/:

Workflow Instruction File Purpose
Conversational Mode workflows/start.md Natural requirements gathering through discussion
Extract Requirements workflows/summarize.md Analyze conversation → mini-PRD + optimized prompts
Prompt Optimization workflows/improve.md Intent detection + quality assessment + auto-depth selection
PRD Generation workflows/prd.md Socratic questions → full PRD + quick PRD
Mode Boundaries core/clavix-mode.md Planning vs implementation distinction
File Operations core/file-operations.md File creation patterns
Verification core/verification.md Post-implementation verification

Read the full file on GitHub · 218 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. 3d ago First seen · 218 lines · 0 tokens per session scan A 7ea23489d78e

Subscribe to this mod's changes

agents is an agent published in the GitHub repository bobvibes/Clavix (673 stars, last pushed 6mo ago), licensed Apache-2.0. It costs nothing until one of its globs matches a file; then it loads 2,055 tokens. 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.