Callable AGENTS.md

Project instructions for Callable, a command layer that lets AI agents discover and run workflows in existing business software through MCP and REST APIs. They require real integrations and real results, with no fake data or mock behavior.

In plain words
What is it for?
Use them when building or changing Callable. They define its product boundaries, require real error handling and audit information, and prohibit mocked workflows, responses, metrics, and success states.
Why use it?
They keep development focused on a narrow, working product and prevent demos from pretending that unavailable connections or actions succeed.

Instructions file for CodexOpenCode

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 instructions/romone6/callable/agents-md
Clone the repo
git clone --depth 1 https://github.com/Romone6/Callable

Made for: Codex, OpenCode.

Per session 649 This file is loaded in full into every session.
When invoked 649 The same file — it is already loaded in full.
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 $0.00649 $0.00649
Opus 5 $0.00324 $0.00324
Sonnet 5 $0.00130 $0.00130
Haiku 4.5 $0.00065 $0.00065

Measured yesterday against content hash e4f9e8e01952, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

Callable AGENTS.md 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 yesterday.

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.

AGENTS.md · 131 lines

How it starts

The opening of the file, as written. The whole thing — 131 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AGENTS.md

Product

Callable is an agent-native command layer for existing business software.

It discovers workflows inside real business tools or real workflow evidence, turns them into structured commands and exposes those commands to AI agents through MCP and REST APIs.

Non-negotiable rule: no mocks

Do not build mocks.

No mock execution engine. No fake command results. No hardcoded workflow discovery. No placeholder integrations labelled as real. No fake dashboard metrics. No static MCP responses. No simulated success states. No fabricated audit logs. No fake drift status.

If a feature is unavailable, mark it as unavailable. If a connection fails, show the real error. If discovery cannot find a workflow, return no workflow found. If execution fails, log the actual failure.

The MVP must be narrow but real.

Correct product framing

Do not build Callable as:

  • a company brain
  • a generic knowledge base
  • a chatbot
  • an RPA clone
  • a document Q&A tool
  • a frontend-only demo

Build Callable as:

  • a command layer for business software
  • a semantic wrapper for existing workflows
  • an MCP/API gateway for agent-executable commands
  • a safe execution system with approvals, audit logs and drift checks

Core MVP flow

The main MVP flow is:

  1. User creates workspace.
  2. User registers a real target app.
  3. User uploads real workflow evidence such as SOPs, CSV exports or traces.
  4. Callable discovers workflow candidates from that evidence.
  5. User accepts a candidate.
  6. Callable generates a command schema.
  7. User reviews and publishes the command.
  8. Agent calls command through MCP/API.
  9. Callable executes command through real API/browser automation.
  10. Execution is logged.
  11. Drift checks verify command health.

Supported first target

Use a real internal controlled target app if external SaaS credentials are unavailable.

This app must have:

  • real database records
  • real forms
  • real API routes
  • real state changes
  • real browser-executable workflows

Read the full file on GitHub · 131 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. yesterday First seen · 131 lines · 649 tokens per session scan A e4f9e8e01952

Subscribe to this mod's changes

Callable AGENTS.md is an instructions file published in the GitHub repository Romone6/Callable (0 stars, last pushed 4d ago), licensed MIT. It adds 649 tokens to every session, about $0.0032 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.