agentleFS
dwsilvardwsilvar/LBaC/.agents/AGENTS.md

LBaC

AGENTS.md0 starsChanged 4 months agoMITView on GitHub

This file contains the development guidelines, architecture conventions, and token optimization rules for AI agents assisting in this repository. MANDATORY FIRST STEP: Before initiating any development task, you must read the Master Blueprint of the project: πŸ“– docs/LBAC.md The Master Blueprint is the Single Source of Truth (SSoT) of this repository. It links to all Feature Maps (As-Is/To-Be), Architectural Decision Records (ADRs), Business Rules, Sprint Schedules, and BDD Acceptance Criteria. Mandatory workflow for any coding task: 1. Read the relevant…

  • Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.

    Of the AGENTS.md files we read, the 31,439 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.

  • In this repository4 other files it gives its agents.

The file

Show its rules and text

We split this AGENTS.md into the rules it gives an agent

13 rules. About 0 are common, 3 are distinctive, and 10 are very specific to this project.

  • β‰ˆ0

    Common Β· estimate

    Many files like this have a rule of this kind.

  • β‰ˆ3

    Distinctive Β· estimate

    This author's own choices. Worth a look.

  • 10

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read the relevant feature map in docs/features/.

    line 15
  2. Very specific:

    Check the technical debt log in docs/technical_debt.md for any related cleanups.

    line 16
  3. Very specific:

    Before coding: Create an Implementation Plan in docs/implementations/ using Template 4 from docs/LBAC.md.

    line 19
  4. Very specific:

    Traceability Updates: When creating or renaming any source file, search all markdown files in docs/ and update their file:/// links to …

    line 24
  5. Very specific:

    Technical Debt Sync: When completing a task that resolves a task listed in docs/technical_debt.md, automatically: (1) move the task from …

    line 25
  6. Very specific:

    Blueprint Index Sync: When creating a new Feature Map, ADR, or Business Rule file, automatically add it to the corresponding index table in …

    line 27
  7. Very specific:

    ADR Status Sync: When a decision is superseded by a new ADR, update the old ADR's Status field to Superseded by ADR-YYY and update its row …

    line 28
  8. Very specific:

    Stale Verification Report Alert: If a Feature Map's "To-Be" section is modified after a Verification Report already exists for that story, …

    line 29
  9. Very specific:

    Gateway File Consistency: When a core workflow rule in this file (.agents/AGENTS.md) changes significantly, propagate the change summary to …

    line 30
  10. Distinctive:

    Localized Edits (No Full Rewrites): When suggesting code changes, never output the entire file.

    line 47
  11. Distinctive:

    Do Not Explain the Obvious: Do not explain language syntax, standard library features, or basic framework architecture concepts unless …

    line 48
  12. Distinctive:

    Surgical Search: Use search and read tools selectively.

    line 49
  13. Very specific:

    No Placeholders: Write complete code segments for implementation.

    line 50

β‰ˆ β€œCommon” is an estimate: at least 1 in 5 of the 40,246 AGENTS.md files we read have a rule of the same kind, not the same words. β€œVery specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
AGENTS.md

πŸ€– Project Rules and AI Agent Context

This file contains the development guidelines, architecture conventions, and token optimization rules for AI agents assisting in this repository.


πŸ“ 0. Living Blueprint-as-Code (LBaC) Framework

MANDATORY FIRST STEP: Before initiating any development task, you must read the Master Blueprint of the project: πŸ“– docs/LBAC.md

The Master Blueprint is the Single Source of Truth (SSoT) of this repository. It links to all Feature Maps (As-Is/To-Be), Architectural Decision Records (ADRs), Business Rules, Sprint Schedules, and BDD Acceptance Criteria.

Mandatory workflow for any coding task:

  1. Read the relevant feature map in docs/features/.
  2. Check the technical debt log in docs/technical_debt.md for any related cleanups.
  3. Respect the technical constraints defined in this file.
  4. Do not create models, screens, classes, or database tables without first verifying if they already exist in the codebase.
  5. Before coding: Create an Implementation Plan in docs/implementations/ using Template 4 from docs/LBAC.md. Present it to the user and wait for explicit approval.
  6. After coding: Complete a Verification Report using Template 5 from docs/LBAC.md, confirming all BDD scenarios from the feature map pass successfully.
  7. Log any technical debt or significant architectural decisions found during development.

πŸ”„ AI Auto-Documentation & Synchronization Rules:

  • Traceability Updates: When creating or renaming any source file, search all markdown files in docs/ and update their file:/// links to match the new paths.
  • Technical Debt Sync: When completing a task that resolves a task listed in docs/technical_debt.md, automatically: (1) move the task from "Pending Tasks" to "Resolved Tasks" with a resolution summary, AND (2) remove or annotate the corresponding entry in the Feature Map's ## πŸ› οΈ Related Technical Debt section.
  • Sprint History Sync: Upon completing a story, update docs/planning/sprint_history.md to mark the story as completed ([x]) and record any key architectural decisions made. When closing a sprint, also update the sprint table in docs/planning/proposal_planning_qa.md.
  • Blueprint Index Sync: When creating a new Feature Map, ADR, or Business Rule file, automatically add it to the corresponding index table in docs/LBAC.md.
  • ADR Status Sync: When a decision is superseded by a new ADR, update the old ADR's Status field to Superseded by ADR-YYY and update its row in the docs/LBAC.md index accordingly.
  • Stale Verification Report Alert: If a Feature Map's "To-Be" section is modified after a Verification Report already exists for that story, add a ⚠️ Requires Re-verification warning to the top of the existing Verification Report in docs/implementations/.
  • Gateway File Consistency: When a core workflow rule in this file (.agents/AGENTS.md) changes significantly, propagate the change summary to the gateway files: CLAUDE.md, .github/copilot-instructions.md, .cursor/rules/lbac_rules.mdc, and .windsurfrules.

🎯 1. Project Purpose & Core Domain

  • Project Name: Living Blueprint-as-Code (LBaC) β€” Template Repository
  • Purpose: A methodology and repository template that turns product specs, architectural decisions, roadmaps, and QA strategies into active, versioned code documents. Acts as the Single Source of Truth (SSoT) for human developers and AI coding agents to align on a project's state without ambiguity.
  • Core Tech Stack: Markdown (documentation), Git (version control). Agnostic by design β€” the framework can wrap any application tech stack (e.g., Node.js/Express, Spring Boot, Flutter, Next.js, PostgreSQL).

⚑ 2. Rules for Token and Credit Optimization (Efficient Consumption)

To save context tokens, minimize API costs, and improve response speed, the AI agent must strictly comply with:

  1. Ultra-Concise Responses: Get straight to the point. Avoid long intros, repetitive greetings, or generic conclusions. Assume the user is an experienced developer.
  2. Localized Edits (No Full Rewrites): When suggesting code changes, never output the entire file. Provide specific code snippets, replacement blocks, or standard unified diffs.
  3. Do Not Explain the Obvious: Do not explain language syntax, standard library features, or basic framework architecture concepts unless explicitly asked.
  4. Surgical Search: Use search and read tools selectively. Read only the line ranges you need.
  5. No Placeholders: Write complete code segments for implementation. Do not use comments like // TODO: implement this later unless specifically directed.

πŸ› οΈ 3. Technical Constraints & Conventions

  • File Linking: In your responses, always format file paths as clickable links using the file:/// scheme and forward slashes (e.g., README.md).
  • Documentation Language Rules:
    • English: docs/LBAC.md, docs/features/*.md, docs/technical_debt.md, docs/planning/sprint_history.md, .agents/AGENTS.md β€” these files are consumed directly by the AI to keep token tokenization efficient (~15-25% savings).
    • Local Language (e.g., EspaΓ±ol): docs/business/*.md, docs/planning/*_qa.md, docs/adr/*.md β€” these files are mostly consumed by human stakeholders.
  • Architecture Flow: Feature Map ──> Implementation Plan ──> Source Code ──> Verification Report ──> Sprint History & Technical Debt Register

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.