agentic-coding-workflow AGENTS.md

A set of instructions for managing feature work in an agentic coding project, where an AI helper assists with development. It requires the work to be understood, split into small chunks, planned in FEATURE_PLAN.md, and reviewed before coding begins.

In plain words
What is it for?
Planning multi-step software changes, organizing work into stacked branches, checking that git-town is installed, and using FEATURE_PLAN.md as the source of truth for each chunk.
Why use it?
It prevents coding from starting before the full scope and an approved plan are clear, and keeps each branch focused on one concern.

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/abhilash-m96/agentic-coding-workflow/agents-md
Clone the repo
git clone --depth 1 https://github.com/abhilash-m96/agentic-coding-workflow

Made for: Codex, OpenCode.

Per session 1,342 This file is loaded in full into every session.
When invoked 1,342 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 $0.01342 $0.01342
Opus 5 $0.00671 $0.00671
Sonnet 5 $0.00268 $0.00268
Haiku 4.5 $0.00134 $0.00134

Measured yesterday against content hash 30078683bb3e, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

agentic-coding-workflow 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 · 186 lines

How it starts

The opening of the file, as written. The whole thing — 186 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Agentic Coding Workflow:

Rules

  1. Never write code before the chunk plan is proposed, reviewed, and approved by the human and persisted to FEATURE_PLAN.md.
  2. One branch = one concern. No exceptions.
  3. Branches must be stacked on each other - not all branching off main.
  4. If a chunk feels too big, split it further.
  5. Verify git-town is installed before starting: git town --version. If not installed, prompt for installing it.
  6. FEATURE_PLAN.md is your source of truth. Re-read it at the start of every chunk.

Step 1 — Understand the Full Scope

  • Read the full feature requirement before doing anything.
  • Identify every change needed end-to-end.
  • Do not write a single line of code until the full scope is clear.

Step 2 — Propose the Chunk Plan

Break the feature into small, logical, ordered layers and present them to the human for review.

Each chunk must:

  • Represent exactly one concern
  • Be independently readable and reviewable
  • Build on top of the previous chunk

Example for a new REST API endpoint:

Order Branch Contains
1 feat/db-schema Schema + migrations only
2 feat/validators Request/response validators only
3 feat/service-layer Business logic only
4 feat/controller Route handlers only

Adapt to the nature of the work. This is an example, not a rigid template. Present your proposed chunks and wait for the human to confirm, adjust, or reject the plan before proceeding.


Step 3 — Write the Finalised Plan

Once the human approves the chunk plan, immediately write it to FEATURE_PLAN.md in the repo root:

# Feature Plan: <feature-name>

## Chunks
- [ ] feat/db-schema — Schema + migrations
- [ ] feat/validators — Request/response validators
- [ ] feat/service-layer — Business logic
- [ ] feat/controller — Route handlers
- [ ] feat/tests — Tests

## Stack Hierarchy
feat/db-schema → main
feat/validators → feat/db-schema
feat/service-layer → feat/validators
feat/controller → feat/service-layer
feat/tests → feat/controller

Read the full file on GitHub · 186 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 · 186 lines · 1,342 tokens per session scan A 30078683bb3e

Subscribe to this mod's changes

agentic-coding-workflow AGENTS.md is an instructions file published in the GitHub repository abhilash-m96/agentic-coding-workflow (2 stars, last pushed 15d ago), licensed MIT. It adds 1,342 tokens to every session, about $0.0067 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.