context-optimization

Codex compatibility note:

• Invoke repository skills with $skill-name in Codex; this mirrored copy rewrites legacy Claude /skill-name references.
• Task tracker mandate: BEFORE executing any workflow or skill step, create/update task tracking for all steps and keep it synchronized as progress changes.
• User-question prompts mean to ask the user directly in Codex.
• Ignore Claude-specific mode-switch instructions when they appear.
• Strict execution contract: when a user explicitly invokes a skill, execute that skill protocol as written.
• Subagent authorization: when a skill is user-invoked or AI-detected and its protocol requires subagents, that skill activation authorizes use of the required spawn_agent subagent(s) for that task.
• Do not skip, reorder, or merge protocol steps unless the user explicitly approves the deviation first.
• For workflow skills, execute each listed child-skill step explicitly and report step-by-step evidence.
• If a required step/tool cannot run in this environment, stop and ask the user before adapting.

Codex Project-Reference Loading (No Hooks)

Codex does not receive Claude hook-based doc injection. When coding, planning, debugging, testing, or reviewing, open project docs explicitly using this routing.

Always read:

• docs/project-config.json (project-specific paths, commands, modules, and workflow/test settings)
• docs/project-reference/docs-index-reference.md (routes to the full docs/project-reference/* catalog)
• docs/project-reference/lessons.md (always-on guardrails and anti-patterns)

Missing/stale context route: If docs/project-config.json, the docs index, lessons.md, CLAUDE.md, AGENTS.md, or any task-required reference doc is missing or stale, auto-run $project-init or the narrow setup route ($project-config, $docs-init, $scan-all, $scan --target=<key>, $claude-md-init) before ordinary project-specific work. If Codex mirrors or AGENTS.md are missing/stale, ask the user to run $sync-codex; do not auto-run it.

Situation-based docs:

• Backend/CQRS/API/domain/entity changes: backend-patterns-reference.md, domain-entities-reference.md, project-structure-reference.md
• Frontend/UI/styling/design-system: frontend-patterns-reference.md, scss-styling-guide.md, design-system/README.md
• Spec authoring, docs/specs/ pathing, or TC format: feature-spec-reference.md, spec-system-reference.md, spec-principles.md
• Behavior/public-contract changes or spec-test-code sync: workflow-spec-test-code-cycle-reference.md plus the spec docs above
• Derived spec indexes/ERDs/reimplementation guides: spec-system-reference.md and source Feature Specs under docs/specs/
• Integration test implementation/review: integration-test-reference.md
• E2E test implementation/review: e2e-test-reference.md
• Code review/audit work: code-review-rules.md plus domain docs above based on changed files

Do not read all docs blindly. Start from docs-index-reference.md, then open only relevant files for the task.

Quick Summary

Goal: Manage context window efficiently to maintain productivity in long Claude Code sessions.

Workflow:

1. Compress — Create context anchors every 10 operations summarizing progress
2. Isolate — Delegate exploration tasks to sub-agents to reduce context usage

Cross-session persistence (saving findings to survive a new session) is out of scope here — use file checkpoints via $checkpoint and restore with $recover (see the memory-management skill).

Key Rules:

• Write context anchor every 10 operations (re-read task, verify alignment, summarize)
• Use offset/limit and grep before reading large files
• Combine search patterns with OR instead of sequential searches
• At 100K tokens: required compression; at 150K: critical save and summarize

Be skeptical. Apply critical thinking, sequential thinking. Every claim needs traced proof, confidence percentages (Idea should be more than 80%).

Context Optimization & Management

Manage context window efficiently to maintain productivity in long sessions.

---

Context Architecture

┌─────────────────────────────────────────────────────────────┐
│                     Context Window (~200K tokens)           │
├─────────────────────────────────────────────────────────────┤
│ System Prompt (CLAUDE.md excerpts)          ~2,000 tokens   │
│ ─────────────────────────────────────────────────────────── │
│ Working Memory (current task state)         ~10,000 tokens  │
│ ─────────────────────────────────────────────────────────── │
│ Retrieved Context (RAG from codebase)       ~20,000 tokens  │
│ ─────────────────────────────────────────────────────────── │
│ Episodic Memory (past session learnings)    ~5,000 tokens   │
│ ─────────────────────────────────────────────────────────── │
│ Tool Descriptions (relevant tools only)     ~3,000 tokens   │
└─────────────────────────────────────────────────────────────┘

---

Two Context Strategies

1. Compressing (Summarize Long Trajectories)

Create context anchors every 10 operations:

=== CONTEXT ANCHOR ===
Current Task: Implement order return request feature
Completed:

- Created Return entity with validation
- Added SaveReturnCommand with handler
- Implemented entity event handler for notifications

Remaining:

- Create GetReturnListQuery
- Add controller endpoint
- Write unit tests

Key Findings:

- Returns use service-specific repository
- Notifications via entity event handlers, not direct calls
- Validation uses validation framework fluent .AndAsync()

# Next Action: Create query handler with GetQueryBuilder pattern

Pre-Compaction Preservation Checklist (canonical for $compact)

Before a manual $compact (or any context compaction), confirm these are saved so they survive the cut — this is the canonical checklist the user-facing $compact alias delegates to:

• Current branch + uncommitted-changes status
• Active file paths being modified
• Any error messages / stack traces (preserve verbatim when mid-bug)
• Key decisions and their rationale
• Pending items from the todo list

Preserve decisions, files modified, current task state. Drop redundant tool outputs, repeated searches, verbose logs. Compact at natural breakpoints (after commits/PR), not mid-task; after compacting, restate the current objective.

2. Isolating (Use Sub-Agents)

Delegate specialized tasks to sub-agents:

// Explore codebase (reduced context)
Task({ agent_type: 'Explore', prompt: 'Find all entity event handlers in the target service' });

// Plan implementation (focused context)
Task({ agent_type: 'Plan', prompt: 'Plan return approval workflow' });

When to Isolate:

• Broad codebase exploration
• Independent research tasks
• Parallel investigations

---

Context Anchor Protocol

Every 10 operations, write a context anchor:

1. Re-read original task from todo list or initial prompt
2. Verify alignment with current work
3. Write anchor summarizing progress
4. Persist to a checkpoint ($checkpoint) if discovering important patterns

=== CONTEXT ANCHOR [10] ===
Task: [Original task description]
Phase: [Current phase number]
Progress: [What's been completed]
Findings: [Key discoveries]
Next: [Specific next step]
Confidence: [High/Medium/Low]
===========================

---

Token-Efficient Patterns

File Reading

// ❌ Reading entire files
Read({ file_path: 'large-file.cs' });

// ✅ Read specific sections
Read({ file_path: 'large-file.cs', offset: 100, limit: 50 });

// ✅ Use grep to find specific content first
Grep({ pattern: 'class SaveOrderCommand', path: '<source-root>/' });

Search Optimization

// ❌ Multiple sequential searches
Grep({ pattern: 'CreateAsync' });
Grep({ pattern: 'UpdateAsync' });
Grep({ pattern: 'DeleteAsync' });

// ✅ Combined pattern
Grep({ pattern: 'CreateAsync|UpdateAsync|DeleteAsync', output_mode: 'files_with_matches' });

Parallel Operations

// ✅ Parallel reads for independent files
[Read({ file_path: 'file1.cs' }), Read({ file_path: 'file2.cs' }), Read({ file_path: 'file3.cs' })];

---

Anti-Patterns

Anti-Pattern	Better Approach
Reading entire large files	Use offset/limit or grep first
Sequential searches	Combine with OR patterns
Repeating same searches	Reuse earlier results
No context anchors	Write anchor every 10 ops
Not using sub-agents	Isolate exploration tasks
Forgetting discoveries	Save findings to a checkpoint

---

Quick Reference

Token Estimation:

• 1 line of code ≈ 10-15 tokens
• 1 page of text ≈ 500 tokens
• Average file ≈ 1,000-3,000 tokens

Context Thresholds:

• 50K tokens: Consider compression
• 100K tokens: Required compression
• 150K tokens: Critical - save and summarize

Related

• memory-management

---

[IMPORTANT] Use task tracking to break ALL work into small tasks BEFORE starting — including tasks for each file read. This prevents context loss from long files. For simple tasks, AI MUST ATTENTION ask user whether to skip.

AI Mistake Prevention — Failure modes to avoid on every task:

Check downstream references before deleting. Deleting components causes documentation and code staleness cascades. Map all referencing files before removal. Verify AI-generated content against actual code. AI hallucinates APIs, class names, and method signatures. Always grep to confirm existence before documenting or referencing. Trace full dependency chain after edits. Changing a definition misses downstream variables and consumers derived from it. Always trace the full chain. Trace ALL code paths when verifying correctness. Confirming code exists is not confirming it executes. Always trace early exits, error branches, and conditional skips — not just happy path. When debugging, ask "whose responsibility?" before fixing. Trace whether bug is in caller (wrong data) or callee (wrong handling). Fix at responsible layer — never patch symptom site. Assume existing values are intentional — ask WHY before changing. Before changing any constant, limit, flag, or pattern: read comments, check git blame, examine surrounding code. Verify ALL affected outputs, not just the first. Changes touching multiple stacks require verifying EVERY output. One green check is not all green checks. Holistic-first debugging — resist nearest-attention trap. When investigating any failure, list EVERY precondition first (config, env vars, DB names, endpoints, DI registrations, data preconditions), then verify each against evidence before forming any code-layer hypothesis. Surgical changes — apply the diff test. Bug fix: every changed line must trace directly to the bug. Don't restyle or improve adjacent code. Enhancement task: implement improvements AND announce them explicitly. Surface ambiguity before coding — don't pick silently. If request has multiple interpretations, present each with effort estimate and ask. Never assume all-records, file-based, or more complex path. Keep domain concepts out of generic/shared/infrastructure layers. A reusable layer (shared library, framework, infra module) must reference NO consumer-specific domain concept — tenant/customer/product IDs, business entities, feature rules. The leak compiles and runs, so it passes review silently while coupling the "reusable" layer to one consumer. Push domain fields/logic down into the consumer via subclass or composition.

Critical Thinking Mindset — Apply critical thinking, sequential thinking. Every claim needs traced proof, confidence >80% to act. Anti-hallucination: Never present guess as fact — cite sources for every claim, admit uncertainty freely, self-check output for errors, cross-reference independently, stay skeptical of own confidence — certainty without evidence root of all hallucination.

MUST ATTENTION apply critical thinking — every claim needs traced proof, confidence >80% to act. Anti-hallucination: never present guess as fact.

MUST ATTENTION apply AI mistake prevention — holistic-first debugging, fix at responsible layer, surface ambiguity before coding, re-read files after compaction.

Closing Reminders

• MANDATORY IMPORTANT MUST ATTENTION break work into small todo tasks using task tracking BEFORE starting
• MANDATORY IMPORTANT MUST ATTENTION search codebase for 3+ similar patterns before creating new code
• MANDATORY IMPORTANT MUST ATTENTION cite file:line evidence for every claim (confidence >80% to act)
• MANDATORY IMPORTANT MUST ATTENTION add a final review todo task to verify work quality

[TASK-PLANNING] Before acting, analyze task scope and systematically break it into small todo tasks and sub-tasks using task tracking.

Hookless Prompt Protocol Mirror (Auto-Synced)

Source: .claude/hooks/lib/prompt-injections.cjs + .claude/.ck.json

[WORKFLOW-EXECUTION-PROTOCOL] [BLOCKING] Workflow Execution Protocol — MANDATORY IMPORTANT MUST CRITICAL. Do not skip for any reason.

Generic portability boundary: Reusable skills and protocol text stay project-neutral; project-specific conventions are discovered from docs/project-config.json and docs/project-reference/. Apply shared AI-SDD from shared/sdd-artifact-contract.md. Read docs/project-config.json and docs/project-reference/docs-index-reference.md, then open the project reference docs named there. For spec, test-case, behavior-change, public-contract, or docs/specs/ work, route through the local spec docs named by the docs index: feature-spec-reference.md, spec-system-reference.md, spec-principles.md, and workflow-spec-test-code-cycle-reference.md when specs/tests/code must stay synchronized. If either file or a required reference doc is missing or stale, auto-run $project-init (or the narrow lower-level route such as $project-config, $docs-init, $scan-all, or $scan --target=<key>) before ordinary project-specific work. Any supported AI tool may execute when this shared context and local docs are available.

1. DETECT: If the prompt starts with an explicit slash skill/workflow command, execute it directly. Otherwise match the prompt against the workflow catalog and skill list.
2. ANALYZE: Choose the best option: execute directly, invoke a skill, activate a standard workflow, or compose a custom step combination.
3. AUTO-SELECT: Pick the best option yourself. Do not ask the user to choose between direct execution, skill, standard workflow, or custom workflow.
4. ACTIVATE: For a selected workflow, call $start-workflow <workflowId>; for a selected skill, invoke that skill; for a custom workflow, sequence custom steps directly; for direct execution, proceed with the task.
5. CREATE TASKS: task tracking for ALL workflow/skill/custom steps before execution when the selected path has multiple steps.
6. EXECUTE: Advance per the Workflow Step Advancement & Parallel Phases rule in your context instructions — model-driven; a sub-agent completion advances a step identically to an inline call; a parallel-phase group is an all-return barrier (advance only after ALL members return, never serialize it) [CRITICAL-THINKING-MINDSET] Apply critical thinking, sequential thinking. Every claim needs traced proof, confidence >80% to act. Anti-hallucination principle: Never present guess as fact — cite sources for every claim, admit uncertainty freely, self-check output for errors, cross-reference independently, stay skeptical of own confidence — certainty without evidence root of all hallucination. AI Attention principle (Primacy-Recency): Put the 3 most critical rules at both top and bottom of long prompts/protocols so instruction adherence survives long context windows. Goal-driven execution: Define success criteria first, loop until verified, and stop only when observable checks pass. Tests verify intent: Tests must protect business rules/invariants and fail when the protected intent breaks, not only mirror current behavior.

[LESSON-LEARNED-REMINDER] [BLOCKING] Task Planning & Continuous Improvement — MANDATORY. Do not skip.

Break work into small tasks (task tracking) before starting. Add final task: "Analyze AI mistakes & lessons learned".

Extract lessons — ROOT CAUSE ONLY, not symptom fixes:

1. Name the FAILURE MODE (reasoning/assumption failure), not symptom — "assumed API existed without reading source" not "used wrong enum value".
2. Generality test: does this failure mode apply to ≥3 contexts/codebases? If not, abstract one level up.
3. Write as a universal rule — strip project-specific names/paths/classes. Useful on any codebase.
4. Consolidate: multiple mistakes sharing one failure mode → ONE lesson.
5. Recurrence gate: "Would this recur in future session WITHOUT this reminder?" — No → skip $learn.
6. Auto-fix gate: "Could $code-review/$code-simplifier/$security-review/$lint catch this?" — Yes → improve review skill instead.
7. BOTH gates pass → ask user to run $learn. [TASK-PLANNING] [MANDATORY] BEFORE executing any workflow or skill step, create/update task tracking for all planned steps, then keep it synchronized as each step starts/completes.