Copilot & AGENTS Optimizer Skill

When to use

Use this skill when:

• Running /docs-optimize to refactor AGENTS.md and copilot-instructions.md
• Adding new content to either file and needing to check if it belongs there
• After changes to .augment/ (new skills, rules, guidelines) that may make content in these files redundant
• When either file exceeds the line budget

Do NOT use when:

• Writing application code
• Creating new skills or commands

Procedure: Optimize copilot/agents files

0. Analyze current state

Before changing anything, read and understand:

1. Read AGENTS.md — current content, line count, what's duplicated vs. unique.
2. Read .github/copilot-instructions.md — same analysis.
3. Scan .augment/ — which skills, rules, and guidelines already cover topics from the files above. 3b. Scan the playbooks too, not only .augment/. A packages/<n>/AGENTS.md may duplicate a playbook rather than a shipped rule, and dedup against .augment/ alone misses it entirely — the content is not in .augment/ at all. Run findRestatedSteps (src/scripts/derive_playbooks.ts) over each workspace file: a line carrying a playbook's invokes id outside a link is a restatement to replace with a pointer. The same id inside a link label is the shape the contract wants — see agents-md-thin-root § Workspace files.
4. Identify duplication — what content exists in both AGENTS.md and .augment/?

Only after this analysis, proceed with optimization.

AGENTS.md — Project Entry Point for AI Agents

Property	Value
Audience	Augment Agent, other AI agents
Can read .augment/?	✅ Yes — can follow references
Size budget	≤ 3,000 chars (package root; consumer template ≤ 2,500) — enforced by lint_agents_md.ts; see size-and-scope.md
Purpose	Project-specific setup, Docker, testing, quality tools

What belongs here:

• Tech stack and framework versions
• Development setup (Docker, Make targets, env files)
• Database and multi-tenancy setup
• Testing framework, suites, and conventions
• Quality tool commands (PHPStan, Rector, ECS)
• Project structure overview (brief, link to module docs)
• Agent infrastructure overview (layer table, key references)

What does NOT belong here:

• Coding standards (→ .augment/rules/ and .augment/guidelines/)
• Architecture principles like SOLID, KISS, DRY (→ .augment/rules/architecture.md)
• PHP conventions (→ ../../../docs/guidelines/php/)
• Scope control rules (→ .augment/rules/scope-control.md)
• Language/tone rules (→ .augment/rules/language-and-tone.md)
• Detailed module documentation (→ app/Modules/README.md)

.github/copilot-instructions.md — Self-Contained for Copilot

Property	Value
Audience	GitHub Copilot (Code Review bot + Chat)
Can read .augment/?	❌ Code Review cannot, ✅ Chat can
Line budget	< 150 lines (ideal < 60); Copilot Code Review reads only the first 4,000 chars
Purpose	Coding standards, review rules, architecture constraints

What belongs here:

• Architecture rules (thin controllers, service layer, policies)
• PHP 8.2 patterns (readonly, final, enums — with exceptions)
• Project-specific conventions (custom helpers, environment config, naming)
• Code review scope and comment behavior rules
• Language rules (English comments, bilingual PR comments)
• Known issues / false positives that Copilot should avoid
• Package management rules

What does NOT belong here:

• Docker setup or Make targets (Copilot doesn't run commands)
• Testing setup details (Copilot doesn't run tests)
• Agent infrastructure (Copilot doesn't use agents/)
• Things auto-enforced by ECS/Rector (code style, formatting, trailing commas)
• Detailed pattern documentation (too long, Copilot needs concise rules)

Deduplication Strategy

Rule: Content lives in ONE canonical place

.augment/rules/         ← Canonical for behavior rules
.augment/guidelines/    ← Canonical for coding conventions
.augment/skills/        ← Canonical for domain expertise
agents/                 ← Canonical for project-specific docs

When duplication is acceptable

copilot-instructions.md must duplicate essential rules because Copilot Code Review cannot read other files. But keep duplicated content:

• Concise — one-liner summaries, not full explanations
• Focused — only what Copilot needs for code suggestions and PR reviews
• Stable — rules that rarely change (architecture, naming, key conventions)

When duplication is NOT acceptable

AGENTS.md should never duplicate .augment/ content because Augment Agent can read both. Instead, reference with a table:

| What | Where |
|---|---|
| PHP coding rules | `.augment/rules/php-coding.md` |
| Controller guidelines | `../../../docs/guidelines/php/controllers.md` |

Line Budget Enforcement

File	🟢 Good	🟡 Warning	🔴 Over budget
AGENTS.md	≤ 2,800 chars	2,801–3,000 chars	> 3,000 chars — lint_agents_md.ts fails (consumer template: warn 2,300 / fail 2,500)
copilot-instructions.md	< 60 lines	60–150 lines	> 150 lines, or key rules past the 4,000-char Copilot read window

Reduction strategies (when over budget)

1. Extract to agents/ — Move project-specific details to dedicated files in agents/ and link from AGENTS.md (e.g., agents/reference/docs/database-setup.md, agents/reference/docs/testing.md)
2. Remove duplicates — If content exists in .augment/, remove from AGENTS.md
3. Condense — Turn verbose explanations into concise tables or bullet points
4. Remove ECS/Rector-enforced rules — From copilot-instructions.md (auto-fixed anyway)
5. Move examples to guidelines — Detailed code examples belong in .augment/guidelines/

Portability and Stack Coherence

Duplication is not the only way these files go wrong. They also lie about the project over time:

• Legacy identifiers — references to a former repo name, a sibling project in the same monorepo, or content that was copied from another codebase as a starting template and never adapted.
• Stack drift — the file claims "Laravel 11 + MariaDB" but the repo was converted to a Python library; or it names Docker services that no longer exist.
• Dead commands — make start in the docs, but Makefile no longer has that target.

Before deduplicating, run three scans:

1. Legacy identifier scan — compare both files against the package's FORBIDDEN_IDENTIFIERS blocklist (see scripts/check_portability.ts) plus any project names from agents/ module docs that don't match the current project.
2. Stack coherence scan — auto-detect the actual stack from composer.json / package.json / pyproject.toml / etc. and flag any claim that no longer matches reality.
3. Dead-command scan — verify every make X, task X, composer X, php artisan X, or npm X still resolves.

Every hit from scan 1 is a 🔴 blocker: leaking another project's name into a consumer's own docs is the failure mode this skill exists to prevent. Fix or remove those BEFORE any dedup/condense work — there's no point deduplicating content that is about to be rewritten.

When the drift is severe (whole sections are wrong), recommend /agents init to scaffold a clean replacement rather than patching forever.

agent-config Path Conventions — Preserve, Don't "Fix"

copilot-instructions.md ships a "Known False Positives" section that tells Copilot Code Review not to flag agent-config path patterns as broken. When optimizing, keep that section intact — never delete it as "redundant" and never trim its bullets. The patterns it covers:

• Relative cross-references inside .augment/ rules / skills (../docs/guidelines/foo.md, ../contexts/bar.md) — paths resolve from the file's delivered location, not from the symlink in .claude/rules/ etc. (per road-to-path-fixes.md Strategy A).
• path_prefix: triggers containing src/ — literal match patterns, not file refs (per Modified Option 1, P2.2).
• Symlinked rule files under .claude/rules/, .cursor/rules/, .clinerules/ — targets resolve into .augment/rules/.

If the consumer project's copilot-instructions.md is missing the section, add it during optimization using the canonical block from .augment/templates/copilot-instructions.md. Surfaces include /agents init and /agents optimize.

Optimization Checklist

When optimizing either file, check:

• "Known False Positives" section present and unmodified?
• No identifiers from other projects (FORBIDDEN_IDENTIFIERS blocklist)?
• Tech stack claims match actual project dependencies?
• All referenced commands/targets exist (Makefile, composer scripts, artisan, task)?
• Line count within budget?
• No content duplicated with .augment/rules/?
• No content duplicated with .augment/guidelines/?
• No content duplicated between the two files unnecessarily?
• AGENTS.md references .augment/ instead of duplicating?
• copilot-instructions.md is self-contained for Code Review?
• No ECS/Rector-enforced rules in copilot-instructions.md?
• Project structure is brief (link to module docs for details)?
• All sections still relevant (no outdated references)?
• Cross-references to agents/ docs are valid?

Related

• Command: /agents optimize
• Skill: copilot-config — Copilot behavior and PR review patterns
• Skill: agent-docs-writing — documentation hierarchy
• Context: augment-infrastructure.md — full .augment/ overview

Output format

1. Optimized file(s) with deduplication applied and line budget respected
2. Summary of what was removed, moved, or consolidated
3. Line count before/after for each file

Gotcha

• AGENTS.md is read by GitHub Copilot and other tools — changes affect all AI assistants, not just Augment.
• Don't remove content from AGENTS.md that other tools depend on — verify cross-tool compatibility first.
• The model tends to over-optimize by removing "obvious" content that other models actually need.

Do NOT

• Do NOT add content to AGENTS.md that belongs in .augment/ files.
• Do NOT exceed the recommended line budget for copilot-instructions.md.
• Do NOT duplicate rules between AGENTS.md and .augment/rules/.

Auto-trigger keywords

• AGENTS.md optimization
• copilot-instructions
• deduplication