Eko GTD - Planning & Execution Methodology

Eko GTD (Eko Getting Things Done) is a structured planning and execution methodology that transforms unstructured requirements into a system of testable challenges with autonomous progress tracking.


Quick Reference

Invocation

Use Eko GTD:<Domain> to <task description>

Domains

DomainUse For
Eko GTD:SoftwareCode, schema, components, APIs
Eko GTD:MarketingCampaigns, metrics, deliverables
Eko GTD:ContentEditorial, SEO, publishing
Eko GTD:VideoScripting, production, post
Eko GTD:AdsDigital advertising campaigns
Eko GTD:ResearchResearch projects, synthesis
Eko GTD:BusinessPlanning, strategy execution
Eko GTD:Doc-UpdatesDocumentation maintenance, post-code-change updates

Resume Work

Continue Eko GTD @path/to/TODO.md

Complexity Tiers

Eko GTD scales from minute tasks to major builds with a 6-tier spectrum:

TierNameDurationTrackingDocument
XSMinute< 30 minTODO.md (flat)Commit message
QSQuick30 min - 2hTODO.md (flat)Commit with checklist
SSmall2-4 hoursTODO.md (mini)challenge-mini.md
MMedium4-6 hoursTODO.md (mini+)Mini + context
LLarge6-10 hoursTODO.md (full)challenge-full.md
XLMajor10+ hoursTODO.md (full)Full + waves

Every tier gets a TODO.md. Even XS/QS tasks create a flat TODO.md for session-resumable context.

Light Tracking (XS/QS)

For XS and QS tasks, skip formal challenges but still create a flat TODO.md. This takes seconds and ensures work is resumable across sessions.

XS (Minute) tasks:

  • Typo fixes, config value tweaks, single-line bug fixes, comment updates
  • Create TODO.md with 2-4 checklist items

QS (Quick) tasks:

  • Single-file changes, straightforward bug fixes, minor refactors, simple UI tweaks
  • Create TODO.md with 4-8 checklist items

Full GTD (S+)

Use formal GTD tracking with challenges when complexity increases:

Invoke GTD when any of these apply:

  • Multiple files across packages
  • Database schema changes
  • API contract changes
  • Cross-domain work (e.g., UI + worker + queue)
  • Cross-agent ownership
  • External service integration
  • Unclear requirements needing exploration

Complexity Detection

Automatic Detection Signals

Claude should assess task complexity using these signals:

SignalXS/QS (Skip GTD)S+ (Invoke GTD)
File count1-2 files3+ files
Package scopeSingle packageCross-package
Schema changesNoneMigrations required
API changesNoneContract changes
DependenciesNoneExternal services
Agent ownershipSingle agentMulti-agent
Estimated duration< 2 hours2+ hours
UncertaintyClear pathExploration needed

Detection Decision Tree

Is this a typo, config tweak, or single-line fix?
├─ YES → XS (just commit)
└─ NO → Is it a single-file change, < 2 hours, no schema/API changes?
         ├─ YES → QS (commit with checklist)
         └─ NO → Does it touch 3+ files OR require schema/API changes?
                  ├─ NO → S (mini challenge)
                  └─ YES → Does it span 2+ packages OR involve external services?
                           ├─ NO → M (mini + context)
                           └─ YES → Is estimated duration < 10 hours?
                                    ├─ YES → L (full challenge)
                                    └─ NO → XL (full + waves)

Agent Behavior

When a user requests a task without explicitly invoking GTD:

  1. Assess complexity using the signals above
  2. XS/QS detected: Create a flat TODO.md in the project directory, then execute directly
  3. S+ detected: Suggest GTD with rationale, e.g.:

    "This task touches 4 files across packages/ui and apps/web, and modifies the API contract. I'll use Eko GTD:Software to structure this work."


Project Storage

Eko GTD projects are stored in docs/projects/<project-name>/:

docs/projects/
├── fix-login-typo/               # XS project (flat TODO only)
│   └── TODO.md                   # Flat checklist
├── user-auth/                    # S+ project (TODO + challenges)
│   ├── TODO.md                   # Progress tracker
│   ├── 01-login-flow.md          # Challenge document
│   ├── 02-registration.md        # Challenge document
│   └── 03-password-reset.md      # Challenge document
├── checkout-redesign/
│   ├── TODO.md
│   └── ...
└── README.md                     # Project index (optional)

Naming conventions:

  • Project directories: kebab-case
  • Challenge files: NN-feature-name.md (numbered for ordering, S+ only)
  • Always include TODO.md — every tier, no exceptions

Exception: The docs/challenges/ directory is reserved for the Eko MVP rollout challenges.


Design Principles

IDPrincipleDescription
GTD-001Challenge-BasedRequirements become numbered PASS/FAIL challenges
GTD-002State TrackingFrontmatter metadata enables session-resumable work
GTD-003Wave OrganizationChallenges grouped by dependency order
GTD-004Quality TiersFunctional + quality challenges for A+ grading
GTD-005Cross-SyncTODO.md files coordinate across domains
GTD-006Agent OwnershipEvery challenge has a single owner agent
GTD-007Scaled StructureFull docs for L/XL, mini for S/M, flat TODO for XS/QS
GTD-008Auto-DetectionComplexity assessed automatically before routing
GTD-009Universal TrackingEvery tier gets a TODO.md for session resumability
GTD-010Gaps Pre-flightS+ challenges validated via /gaps before execution begins

Four-Phase Workflow

PHASE 0: GAPS PRE-FLIGHT (S+ tiers)

After creating challenges but before starting implementation, run a /gaps analysis on the challenge set. This catches plan-level problems — missing requirements, contradictory criteria, fragile assumptions — before they become expensive rework.

  1. Analyze — Run /gaps against challenges + TODO.md across six dimensions:
    • Holes, Gaps, Pitfalls, Inconsistencies, Weaknesses, Untapped Opportunities
  2. Record — Write findings to ## Gaps Review section of TODO.md
  3. Triage — Classify each finding:
    • 🔴 Block: Must fix before Phase 1 (amend challenges or add new ones)
    • 🟡 Track: Note for awareness during implementation
    • 🟢 Defer: Nice-to-have, capture for future consideration
  4. Gate — Phase 1 begins only when all 🔴 items are resolved

Skip conditions: XS/QS tiers skip. S tiers may use a lightweight single-pass review.

PHASE 1: CHALLENGES (Implementation Driver)

  1. Create Challenge Documents

    • Decompose source/requirements into challenge docs
    • Assign complexity (S/M/L/XL) and owner agents
    • Define dependencies (wave structure)
  2. Evaluate Challenges

    • Assess each challenge against current state
    • Mark PASS (exists) or FAIL (needs work)
    • Update frontmatter counts
  3. Implement by Wave

    • Work through waves in dependency order
    • Complete each challenge to PASS
    • Re-evaluate until all PASS
    • Update TODO.md with <- CURRENT marker

PHASE 2: DOCUMENTATION

  1. Update specs/docs to reflect implemented features
  2. Ensure all TODO files are consistent
  3. Verify links between documents

PHASE 3: LAUNCH PREP

  1. Polish, edge cases, optimization
  2. Final quality verification
  3. Archive completed challenges and ship

Claude Behavior on Invocation

When user says "Use Eko GTD: to...":

  1. Confirm Scope - Ask clarifying questions about requirements
  2. Assess Complexity - Determine tier (XS/QS/S/M/L/XL)
  3. Create TODO.md - Every tier gets a TODO.md. Use flat template for XS/QS, mini or full for S+
  4. Propose Structure (S+ only) - Suggest challenge count and wave organization
  5. Create Challenge Docs (S+ only) - Generate challenge documents
  6. Run Gaps Pre-flight (S+ only) - Execute /gaps on challenges, record in TODO.md, triage as 🔴/🟡/🟢, resolve blockers
  7. Begin Phase 1 - Start evaluating challenges (or work through checklist for XS/QS)

Document Templates

Flat TODO.md (XS/QS)

---
title: "[Task Name]"
tier: XS | QS
status: active
last_updated: YYYY-MM-DD
---

# [Task Name]

## Goal

[One sentence: what does "done" look like?]

## Checklist

- [ ] Step 1
- [ ] Step 2
- [ ] Step 3

## Files

| File | Change |
|------|--------|
| `path/to/file.ts` | [Brief description] |

## Notes

[Context, decisions, gotchas — append as you work]

Mini Challenge (S/M)

---
title: "Challenge: [Feature]"
status: in_progress
complexity: S
estimated_challenges: 5
pass_count: 0
fail_count: 0
owner: [agent]
---

# [Feature]

## Context
Brief description of the requirement.

## Challenges

### Challenge 1.1: [Name]
**Requirement:** [Statement]
**Acceptance Criteria:**
- [ ] Criterion 1
- [ ] Criterion 2
**Evaluation:** PENDING
**Notes:**

## Quality Tier

### Challenge 1.X: Quality Standards
**Requirement:** Meet domain-specific quality criteria.
**Acceptance Criteria:**
- [ ] Quality criterion 1
**Evaluation:** PENDING

Full Challenge (L/XL)

---
title: "Challenge: [Feature]"
status: draft|in_progress|complete
complexity: L|XL
estimated_challenges: N
pass_count: X
fail_count: Y
pending_count: Z
dependencies: [list]
owner: [agent]
evaluated_at: YYYY-MM-DD
target_grade: "A+ (97+)"
achieved_grade: null
---

# [Feature]

## Context
What the source/requirements describe.

## Current State
What exists today. Key files, endpoints, components.

## Gap Analysis
| Feature | Required | Current | Gap |
|---------|----------|---------|-----|
| Feature 1 | Yes | Missing | FAIL |

## Challenges

### Challenge X.Y: [Name]
**Requirement:** Statement of what must be true.
**Acceptance Criteria:**
- [ ] Criterion 1
- [ ] Criterion 2
**Evaluation:** [x] PASS | [ ] FAIL | [ ] PENDING
**Notes:** Implementation details.

---

## Quality Tier Challenges (A+ Grade)

### Challenge X.Y: [Quality Area]
**Requirement:** Domain-specific quality standard.
**Acceptance Criteria:**
- [ ] Quality criterion 1
**Evaluation:** PENDING

---

## Evaluation Summary

| Challenge | Result |
|-----------|--------|
| X.1 Name | PASS |
| X.2 Name | FAIL |

**Score: X/Y PASS**

---

## Schema Impact (Software)
Database/API changes required.

## UI Components (Software)
New components to create.

## Implementation Notes
Key decisions and rationale.

## Related Documents
Cross-references.

TODO.md

---
title: "[Project] Progress Tracker"
current_phase: "PHASE 0: GAPS PRE-FLIGHT"
current_wave: "Wave 1: [Name]"
current_task: "01-challenge.md"
last_updated: YYYY-MM-DD
blocked_by: null
---

# [Project] Progress

## User Instructions

### Preferences
- [Workflow preferences]

### Blockers to Address
- [Known blockers]

### Skip/Defer
- [Items to skip]

### Priority Overrides
- [Priority notes]

---

## Quality Reference
- **Target Grade:** A+ (97+)
- **Key Metrics:** [Domain-specific metrics]

---

## Gaps Review

> Generated by `/gaps` pre-flight analysis. Re-run: `/gaps @path/to/TODO.md`

| # | Dimension | Finding | Severity | Resolution |
|---|-----------|---------|----------|------------|
| | | _Run `/gaps` to populate_ | | |

**Blockers:** 0 🔴 | **Tracked:** 0 🟡 | **Deferred:** 0 🟢

---

## Phase Transition Checklists

### PHASE 0 -> PHASE 1
- [ ] `/gaps` analysis completed
- [ ] All 🔴 findings resolved
- [ ] 🟡 findings noted in relevant challenge docs

### PHASE 1 -> PHASE 2
- [ ] All wave challenges PASS
- [ ] No blockers
- [ ] Quality tier verified

### PHASE 2 -> PHASE 3
- [ ] All docs updated
- [ ] Cross-refs verified

---

## Wave 1: [Name]

### 01-challenge.md <- CURRENT
- [x] Create challenge document
- [ ] Evaluate challenges
- [ ] Implement
- [ ] Re-evaluate until PASS

---

## Blocked
<!-- Items waiting -->

## Completed
### [Document] (YYYY-MM-DD)
- **Grade:** A+
- **Deliverables:** [List]

---

## Workflow Reference

### Commands
@path/to/TODO.md What's the current task?
@path/to/TODO.md Resume from where we left off
@path/to/TODO.md Mark [challenge] complete
@path/to/TODO.md Ready for Phase N?
@path/to/TODO.md Block on [reason]
/gaps @path/to/TODO.md               # Re-run gaps analysis

Domain-Specific Quality Tiers

Eko GTD:Software

CategoryStandard
AccessibilityWCAG 2.1 AA compliance
MotionCLS=0, 60fps, prefers-reduced-motion
TestingUnit, integration, Storybook stories
TokensAll styling uses design tokens

Eko GTD:Marketing

CategoryStandard
BrandVisual identity alignment
TargetingAudience segment precision
MetricsConversion tracking defined
ComplianceLegal, platform policies

Eko GTD:Content

CategoryStandard
VoiceTone consistency score
SEOKeyword optimization
ReadabilityFlesch-Kincaid score
AccuracyFact-check verification

Eko GTD:Video

CategoryStandard
ScriptFlow and pacing review
VisualShot composition check
AudioClarity verification
A11yCaptions, descriptions

Eko GTD:Ads

CategoryStandard
CreativeCTR benchmarks met, A/B variants defined
AudienceSegment overlap < 15%, exclusion lists applied
BudgetROAS target set, daily cap enforced
PolicyPlatform ad policies verified per network

Eko GTD:Business

CategoryStandard
FinancialsAssumptions explicit, projections sensitivity-tested
AlignmentStakeholder sign-off criteria defined
RiskTop 3 risks identified, mitigations assigned
TimelineMilestones dated, dependencies mapped

Eko GTD:Doc-Updates

CategoryStandard
AccuracyAssertions match current code behavior
Cross-refsInternal links valid, index entries current
ChangelogEntry added per CHG-001 if functional change
ConventionsFrontmatter, naming, structure per CONVENTIONS.md

Eko GTD:Research

CategoryStandard
MethodologyApproach documented, bias controls stated
SourcesPrimary sources cited, recency verified
SynthesisClaims traceable to evidence, gaps flagged
ReproducibilitySteps repeatable, data access documented

Agent Ownership Rules

IDRuleDescription
OWN-001Single ownerEach challenge doc has exactly one owner
OWN-002Domain-basedOwner matches primary domain of work
OWN-003DelegationComplex work may delegate but owner validates
OWN-004PASS authorityOwner confirms PASS criteria before completion

Workflow Commands

CommandPurpose
@path/to/TODO.md What's the current task?Show current work
@path/to/TODO.md ResumeContinue from last position
@path/to/TODO.md Mark X completeUpdate task status
@path/to/TODO.md Ready for Phase N?Check transition gate
@path/to/TODO.md Block on [reason]Add blocker
/gaps @path/to/TODO.mdRe-run gaps analysis on challenge set