EXECUTION CHECKLIST: MVP → V2.1 DELIVERY

Format: Implementation roadmap (realized step-by-step)
Date: January 12, 2026
Status: Ready to execute

PHASE 1: V2.0.0-RC1 (MVP RELEASE) - 2-3 WEEKS

MVP-1: CI Green / Test Stability (2-4h) ✅ COMPLETE

What: Fix flaky timeout test, make CI consistent

Steps:

Identify flaky test file: test/integration/filediscovery-real-git.test.ts (git timeout issue)
Conditional timeout: timeout = process.env.CI ? 20000 : 10000
Add CI check in jest.config.cjs
Run locally: npm test → 1109 passed, 0 failures ✅
Commit: "MVP-1: CI green - conditional timeouts for git operations"
Push: git push origin mvp-1-ci-green

Tests:

All 1109 tests passing locally (5-10s timeout)
Configuration verified: CI=true gets 20s timeout

DoD: ✅ CI green, 0 flaky tests
PR: "MVP-1: CI green - conditional timeouts"
Push: git push origin mvp-1-ci-green ✅ DONE

Changes:

jest.config.cjs: Added testTimeout: process.env.CI ? 20000 : 10000
test/integration/filediscovery-real-git.test.ts:
- Detached HEAD tests: process.env.CI ? 15000 : 10000
- Performance test: process.env.CI ? 30000 : 15000

MVP-3: Cerber Doctor Command (3h) ✅ COMPLETE

What: Diagnostic CLI tool that shows setup status

Function: src/cli/doctor.ts

✅ Tool detection: actionlint, zizmor, gitleaks
✅ Contract scan: CERBER.md existence check
✅ Version extraction for installed tools
✅ Install command hints for missing tools
✅ Exit codes: 0 (all OK), 1 (warnings), 2 (blockers)
✅ Output in <1s (actual: ~84-124ms)

Tests: test/cli/doctor.test.ts ✅ 19/19 PASSING

✅ Tool detection tests (actionlint, zizmor, gitleaks)
✅ Version extraction tests
✅ Install command generation
✅ Exit code validation (0/1/2)
✅ Performance tests (<1s)
✅ Integration tests

Example Output:

[Cerber Doctor] Setup Validation

✅ Contract: .cerber/contract.yml found
   Profile: solo

Tools:
✅ actionlint v1.6.27
⚠️  zizmor not found
✅ gitleaks v8.18.0

Status: 2/3 tools ready

Fix: npm install zizmor

DoD: ✅ npx cerber doctor <1s, user knows "what to do next"
PR: "MVP-3: cerber doctor diagnostic command"
Push: ✅ Pushed to mvp-1-ci-green

Implementation Details:

getDoctorToolStatus(): Detects installed tools and versions
runDoctor(): Aggregates contract + tool status
printDoctorReport(): Formats output with actionable hints
All 19 tests pass, 0 failures

MVP-4: Guardian Fast Mode <2s (3-5h) ✅ COMPLETE

What: Pre-commit hook that doesn't block commits (fast staged file detection)

Function: src/cli/guardian.ts ✅

✅ Staged file detection via git diff --cached --name-only
✅ Smart filtering: only .github/workflows/* and .cerber/* files matter
✅ Fast exit (50ms) when no relevant files changed
✅ Performance: <2s on normal commit, <500ms when nothing relevant
✅ Debug mode shows execution timing
✅ Exit codes: 0 (pass), 1+ (violations)

Tests: test/cli/guardian.test.ts ✅ 16/16 PASSING

✅ Staged files detection tests (empty, staged-only, filters)
✅ Guardian execution tests (staged flag, fast path)
✅ Performance tests (<2s, <500ms no-relevant)
✅ Exit code validation
✅ Output formatting with debug flag
✅ Integration test (pre-commit simulation)

Example Usage:

# Fast pre-commit hook
$ npx guardian --staged
✅ OK (no relevant files changed)

# With debug timing
$ npx guardian --staged --debug
[Guardian] No relevant files changed. Exit in 47ms

# When workflows changed
$ npx guardian --staged
[Guardian] actionlint: violations found
exit code 1

DoD: ✅ <500ms when nothing relevant, <2s on normal commit
PR: "MVP-4: guardian fast pre-commit hook"
Push: ✅ Pushed to mvp-1-ci-green

Implementation Details:

getStagedFiles(): Lists files via git diff --cached
filterRelevantFiles(): Only .github/workflows/* and .cerber/*
runGuardian(): Executes checks, returns timing + exit code
All 16 tests pass, covers staged detection & performance

MVP-2: Resilience Wiring (2-3h) ✅ COMPLETE

What: Connect existing resilience code to Orchestrator - ALREADY DONE

Status: ✅ 100% Complete (verified)

✅ src/core/Orchestrator.ts: Uses AdapterExecutionStrategy
✅ Profile-based selection: LegacyExecutionStrategy vs ResilientExecutionStrategy
✅ Strategy pattern properly implemented with dependency injection
✅ Resilience checks in: executeValidation() and executeIntegration()

Tests: 84 PASSING (Full resilience suite)

✅ test/core/resilience/resilience-coordinator.test.ts (13.9s)
✅ test/core/resilience-factory.test.ts
✅ test/core/strategies/resilient-execution-strategy.test.ts
✅ test/core/strategies/legacy-execution-strategy.test.ts
✅ test/core/strategies/adapter-executor.test.ts
✅ test/core/retry-strategy.test.ts
✅ Circuit breaker state transitions tested
✅ Retry strategies tested (exponential, linear, fibonacci)
✅ Coordinator resilience coordination tested
✅ No performance degradation vs legacy (<5% overhead verified)

Verified Features:

✅ Adapter fails → resilience retries (exponential backoff)
✅ Adapter fails N times → circuit breaker opens
✅ Circuit breaker HALF_OPEN → allows recovery attempt
✅ Profile selection: solo → legacy, team → resilient
✅ Orchestrator delegates to strategy (DIP pattern)
✅ All 84 tests green, 0 failures

DoD: ✅ Resilience works in main flow, tested, <5% overhead
PR: "MVP-2: resilience wiring" (already merged) Status: Already integrated into main codebase

MVP-5: Documentation Cleanup (1-2h)

What: Archive old roadmaps, keep ONE_TRUTH_MVP as reference

Steps:

Verify ONE_TRUTH_MVP.md exists and links correctly
Add archive headers to 8 old files (already done)
Verify AGENTS.md has 10 rules (already done)
Update README.md: "See ONE_TRUTH_MVP.md for development roadmap"
Final check: no new docs created, only archive markers

DoD: One source of truth, no conflicting docs
PR: "MVP-5: documentation cleanup (archive old roadmaps)"
Push: git push origin MVP-5-docs

RC1 RELEASE

Checklist before tag:

All 5 MVP routes merged to main
npm test → 1109 passing, 0 failures (local)
CI green: GitHub Actions all checks pass
No new linter warnings: npm run lint ✅
Build works: npm run build ✅
Version updated: package.json → 2.0.0-rc1
CHANGELOG.md updated (brief, no essays)
README.md updated: quick-start working

Tag & Release:

git tag v2.0.0-rc1
npm publish (if desired)
git push origin v2.0.0-rc1
# Create GitHub Release page with CHANGELOG excerpt

RC1 Announcement:

v2.0.0-rc1: MVP Release
- cerber init ✅
- cerber validate ✅  
- cerber doctor ✅ (new)
- guardian <2s ✅ (fast pre-commit)
- 1109 tests (0 failures)

Ready for feedback. Resilience & observability in v2.0.0 full.

PHASE 2: V2.0.0 (FULL RELEASE) - AFTER RC1 FEEDBACK

⚠️ PRE-PHASE-2 SHORTCUTS REMEDIATION:

Before starting Phase 2 features, fix 10 identified shortcuts from MVP:

Add error boundary + logging to doctor/guardian
Implement tool config system (not hardcoded)
Integrate ResilientExecutionStrategy (instead of raw execSync)
Structure stderr error reporting
Make timeouts configurable
Improve test mocks (test failure paths)
Add observability integration (logs, metrics, traces)
Implement performance profiling (P99, not just <1s)
Safe process output parsing with schema validation
Standardize contract format (deprecate CERBER.md)

Effort: 17.5h total (12.5h critical path + 5h quality)
Target: Complete before GitHub PR review

See SHORTCUTS_AUDIT.md for detailed fixes.

V2.0.0-Full-1: Resilience Integration (Hardened from MVP-2)

What: Harden MVP-2 resilience + fix guardian/doctor integration issues

Shortcuts to Fix (from SHORTCUTS_AUDIT.md):

Guardian integration: Use ResilientExecutionStrategy instead of raw execSync
Error recovery: Add retry logic with exponential backoff
Circuit breaker: Integrate for cascading failure prevention
Test scenarios: Add real failure simulation (timeout, crash, bad output)

Implementation:

Refactor guardian execSync → Orchestrator strategy delegation
Add resilience coordinator to doctor tool detection
Implement exponential backoff (1s, 2s, 4s, 8s, stop)
Circuit breaker: 3 failures → OPEN → 30s TTL → HALF_OPEN
Fallback for tools: gracefully degrade if tool unavailable

Tests: Extend test/integration/resilience-full.test.ts

Doctor continues if one tool fails (resilience)
Guardian retries actionlint on transient failure
Circuit breaker prevents tool spam
Cascade failure handled properly

DoD: Doctor/Guardian use resilience patterns, all failure modes tested
PR: "refactor: harden doctor/guardian with resilience (v2.0.0)"
Push: git push origin v2.0.0-resilience-hardened

V2.0.0-Full-2: Observability + Tool Configuration (8h)

What: Logger + metrics wired to doctor/guardian + config-driven tools

Shortcuts to Fix (from SHORTCUTS_AUDIT.md #1, #2, #4, #7, #8):

Add structured logging to doctor/guardian
Implement tool config loader (tools.yml instead of hardcoded)
Parse + structure stderr output
Integrate Logger + metrics collection
Add performance profiling (P99, not just <1s)

Implementation:

Create src/config/tools-config.yml with tool definitions + timeouts
Implement tool config loader with schema validation
Wire logger to doctor/guardian execution paths
Add metrics collection:
- Tool detection success rate
- Execution time (P50/P95/P99)
- Tool output parse success rate
- Error counts by type
Structured stderr parsing per tool
Execution context/span IDs for tracing
Performance profiler (breakdown per tool)

Tests: test/integration/observability.test.ts

Logs emitted for each doctor/guardian step
Metrics incremented correctly
Span IDs track execution flow
P99 latency measured and reported
Tool config loading works (defaults + overrides)
Parse error handling with detailed messages

DoD: Doctor/Guardian observable + configurable, <2% overhead
PR: "feat: observability + tool config (v2.0.0)"
Push: git push origin v2.0.0-observability-config

V2.0.0-Full-3: Error Handling + Output Parsing (5h)

What: Robust error handling, timeout configuration, safe output parsing

Shortcuts to Fix (from SHORTCUTS_AUDIT.md #3, #5, #9):

Make timeouts configurable (not hardcoded 5000ms)
Implement tool-specific output parsers with validation
Add graceful timeout + failure degradation
Support fallback strategies

Implementation:

Load timeouts from contract + tool config
Add per-tool timeout settings in tools.yml
Implement tool-specific output parsers:
- actionlint version parser + schema
- zizmor version parser + schema
- gitleaks version parser + schema
Graceful timeout handling (report separately, don't crash)
Support skip-on-timeout mode for optional tools
Version range validation (SemVer)
Validate parsed output against schema
Report parse errors with suggestions

Tests: test/cli/doctor-parsing.test.ts, test/cli/guardian-errors.test.ts

Test each tool parser with real/fake output
Test timeout behavior (graceful degrade)
Test version range validation
Test parse error messages (helpful)
Test schema validation failures
Test fallback strategies

DoD: All error paths tested, timeouts configurable, parsing safe
PR: "feat: robust error handling + output parsing (v2.0.0)"
Push: git push origin v2.0.0-error-handling

V2.0.0-Full-4: Test Coverage Hardening (5h)

What: Comprehensive test scenarios covering all failure modes

Shortcuts to Fix (from SHORTCUTS_AUDIT.md #6):

Implementation:

Create tool mock factory with states:
- NOT_INSTALLED
- SUCCESS
- TIMEOUT
- INVALID_OUTPUT
- PERMISSION_ERROR
- CRASH
- SLOW (near timeout)
Mock execSync behavior per scenario
Extend doctor.test.ts with failure paths
Extend guardian.test.ts with failure paths
Test graceful fallback behavior
Test error message quality (helpful to user)

Tests: test/cli/doctor-failure-modes.test.ts, test/cli/guardian-failure-modes.test.ts

Test each tool mock state
Test cascade failures (2+ tools fail)
Test recovery from transient failures
Test fast-fail scenarios
Test output validation (safe parsing)
Coverage target: >95% for failure paths

DoD: All failure modes tested, 95%+ coverage, doctor/guardian robust
PR: "test: comprehensive failure mode coverage (v2.0.0)"
Push: git push origin v2.0.0-test-hardening

V2.0.0-Full-5: State Machine / ExecutionContext (8h)

What: Full state tracking for multi-step execution, debug/recovery

Existing Code: State machine 40% complete, needs integration

Steps:

Complete state machine: IDLE → LOADING → EXECUTING → REPORTING → DONE
ExecutionContext tracks: current state, progress, errors, recovery path
Profile + contract determine state transitions
Debug mode: state dumps at each step
Recovery: save/restore execution state for resumable flows
Test: state transitions valid, no orphaned states

Tests: test/integration/state-machine.test.ts

Test state transitions follow contract
Test invalid transitions rejected
Test debug output shows state + context
Test recovery restores from saved state
Test profile "team" requires full state tracking

DoD: State machine operational for multi-step workflows
PR: "feat: state machine & execution context (v2.0.0)"
Push: git push origin v2.0.0-state-machine

V2.0.0 Full Release

Checklist before GA tag:

All 4 shortcuts remediation PRs merged (resilience, observability, error-handling, test-hardening)
npm test → 1200+ passing, 0 failures
Performance: P99 latency <100ms for doctor, <500ms for guardian
Security: no new vulnerabilities
Coverage: >95% for doctor/guardian failure paths
README: updated with new diagnostic features
CHANGELOG: v2.0.0 summary with shortcuts fixed

Tag:

git tag v2.0.0
npm publish
git push origin v2.0.0

v2.0.0 Announcement:

v2.0.0: Production-Ready Release
✅ Resilience: Adaptive strategies with circuit breaker
✅ Observability: Structured logging + metrics
✅ Doctor Diagnostic: Setup health check
✅ Guardian <500ms: Fast pre-commit hook
✅ Error Handling: Robust with graceful degradation
✅ Config-Driven: Tools configurable, not hardcoded

1200+ tests. Enterprise-ready.
Shortcuts from MVP remediated.

PHASE 3: V2.0.1 (PATCH RELEASE)

V2.0.1: Bug Fixes + Micro-optimizations

What: User feedback from rc1 + v2.0.0, quick fixes

Steps:

Collect issues reported in v2.0.0-rc1 period
Priority: blocking bugs only (fix immediately)
Performance: <2s target for guardian (if not met in MVP-4)
Edge cases: Windows path handling, CI timeouts
Each fix: test + DoD + push

Bug Template:

[ ] Bug: [issue name]
    Steps: [reproduce]
    Fix: [code change location]
    Test: [new test file]
    Verify: [benchmark/log output]
    PR: "fix: [issue] (v2.0.1)"

DoD: 0 regression, all new tests pass
Tag: v2.0.1

PHASE 4: V2.1 (FEATURE RELEASE) - 2-3 MONTHS AFTER V2.0

V2.1-1: Auto-install Missing Tools (6h)

What: Cerber detects missing tools and offers to install them

Steps:

Implement tool installer: src/tools/installer.ts
Detection: tool not in PATH
Options:
- --auto-install: download to ~/.cerber/tools/
- --no-install: show hint (current behavior)
- Default: show hint + offer --auto-install
Download from GitHub releases (pinned versions)
Cache in ~/.cerber/tools/ with checksum validation
Fallback: if auto-install fails, fall back to hint

Tests: test/tools/installer.test.ts

Test tool detection (installed vs missing)
Test download + checksum validation
Test cache hit (no re-download)
Test cache miss (re-download)
Test fallback when network unavailable

Integration:

cerber validate --auto-install     # Installs missing tools to ~/.cerber/tools
cerber guardian --auto-install     # Guardian uses installed tools

DoD: Auto-install transparent, all tools downloadable
PR: "feat: auto-install missing tools (v2.1)"
Push: git push origin v2.1-auto-install

V2.1-2: Execution Persistence (6h)

What: Save execution results, allow resumable validation

Steps:

Implement persistence: src/persistence/store.ts
Store: execution ID, timestamp, results, state
Location: .cerber/.cache/executions/<id>.json
Query: cerber report --id <execution-id>
Resume: cerber validate --resume <execution-id> (skip unchanged files)
TTL: cleanup old executions (7 days default)

Tests: test/persistence/store.test.ts

Test execution saved with ID
Test results retrievable by ID
Test resume skips unchanged files
Test TTL cleanup removes old entries
Test concurrent executions don't interfere

Integration:

EXEC_ID=$(cerber validate --solo 2>&1 | grep "Execution ID")
cerber report --id $EXEC_ID                    # Show results
cerber validate --resume $EXEC_ID --solo      # Re-run changed files only

DoD: Execution history available, resume works
PR: "feat: execution persistence (v2.1)"
Push: git push origin v2.1-persistence

V2.1-3: SARIF Format Output (4h)

What: Export results in SARIF 2.1.0 format (GitHub/IDE integration)

Steps:

Implement SARIF exporter: src/reporting/sarif.ts
Map violations → SARIF rules + results
Support: GitHub uploaded artifact, local file
Severity mapping: error/warning/note
Schema: validate against SARIF 2.1.0 spec
Test: output parseable by GitHub/IDE tools

Tests: test/reporting/sarif.test.ts

Test violation → SARIF result conversion
Test severity mapping correct
Test output validates against schema
Test GitHub upload format (if applicable)

Integration:

cerber validate --format sarif > results.sarif
# Upload to GitHub artifact or IDE

DoD: SARIF output validated, GitHub-compatible
PR: "feat: SARIF format output (v2.1)"
Push: git push origin v2.1-sarif

V2.1-4: Plugin System (8h)

What: Allow external adapters via plugin interface

Steps:

Define plugin interface: src/plugin/PluginInterface.ts
Plugin discovery: ~/.cerber/plugins/ directory
Plugin loading: dynamic import with safety
Plugin validation: must implement Adapter interface
Plugin lifecycle: init → run → cleanup
Contract support: plugins in tool list

Plugin Interface:

interface CerberPlugin {
  name: string;
  version: string;
  isInstalled(): Promise<boolean>;
  run(options): Promise<AdapterResult>;
  parseOutput(raw: string): Violation[];
}

Tests: test/plugin/loader.test.ts

Test plugin discovery from directory
Test plugin validation (must implement interface)
Test plugin isolation (no access to internal state)
Test plugin error handling (doesn't crash core)
Test plugin in profile configuration

Integration:

# User creates ~/.cerber/plugins/my-checker/index.ts
cerber validate --solo              # Loads my-checker automatically
# Results include my-checker findings

DoD: Plugin system operational, safe plugin loading
PR: "feat: plugin system (v2.1)"
Push: git push origin v2.1-plugin-system

V2.1-5: Advanced Profiles (3h)

What: Conditional profiles based on environment/context

Steps:

Extend profile syntax: if/then/else conditions
Conditions: env vars, file existence, CI detection
Example: if CI=true then team-strict else dev-fast
Contract support: profile selection via condition
Test: conditions evaluate correctly

Profile Example:

profiles:
  auto:
    if: $CI == "true"
      then: team-strict
    else:
      if: $USER_TYPE == "contractor"
        then: team-limited
      else: dev-fast

Tests: test/contract/profiles-advanced.test.ts

Test condition evaluation
Test nested conditions
Test env var substitution
Test CI detection (GitHub Actions, CircleCI, etc)

DoD: Advanced profiles work in contract
PR: "feat: advanced profile conditions (v2.1)"
Push: git push origin v2.1-advanced-profiles

V2.1 Release

Checklist:

All 5 V2.1 features merged
npm test → 1150+ passing (new tests for each feature)
Performance: no regression vs v2.0.0
Docs: each feature has usage example
README: updated with new features

Tag:

git tag v2.1.0
npm publish
git push origin v2.1.0

v2.1 Announcement:

v2.1.0: Advanced Features
✅ Auto-install tools: --auto-install flag
✅ Execution persistence: Save & resume runs
✅ SARIF export: GitHub/IDE integration
✅ Plugin system: Custom adapters
✅ Advanced profiles: Conditional configuration

1150+ tests. Enterprise-ready.

OVERALL METRICS

Phase	Version	Effort	Timeline	Tests	Status
1	rc1	18-32h	2-3 weeks	1109	MVP ready
1.5	v2.0.0-rc1 fixes	17.5h	2 weeks	1200+	Shortcuts remediated
2	v2.0.0	25h	3 weeks after rc1	1200+	Production ready
3	v2.0.1	TBD	1-2 weeks	TBD	Patch fixes
4	v2.1	27h	2-3 months	1300+	Advanced features

KEY CHANGES FROM MVP AUDIT

Shortcuts Found: 10
Effort to Fix: 17.5h before Phase 2
Impact: High quality, production-ready code
Status: Ready for remediation in Phase 2

What Changed:

MVP-2.0.0-Full-1: Now includes doctor/guardian resilience hardening
MVP-2.0.0-Full-2: Combines observability + tool config (single pass)
MVP-2.0.0-Full-3: Adds error handling + output parsing safety
MVP-2.0.0-Full-4: Adds test hardening for all failure modes
MVP-2.0.0-Full-5: State machine as final v2.0.0 feature

Tests Growth:

MVP: 1109 tests
After shortcuts fix: 1200+ tests (91 new test cases)
After Phase 2: 1300+ tests

EXECUTION RULES

Every feature gets tests FIRST - Write test, make it fail, write code, make it pass
Production code only - No shortcuts, no hacks, no "TODO" comments
Senior dev standard - Code review by senior, all PRs require DoD
Push each route - Don't batch, push MVP-1, then MVP-3, then MVP-4, etc.
No documentation - Only update ONE_TRUTH_MVP.md with completed routes
CI must be green - Every push must pass all tests on GitHub Actions
Performance first - Benchmark every change, document overhead
Backward compatible - No breaking changes until major version

This is the execution plan. Follow it sequentially. No deviations.

Navigation Menu

Uh oh!

EXECUTION CHECKLIST: MVP → V2.1 DELIVERY

PHASE 1: V2.0.0-RC1 (MVP RELEASE) - 2-3 WEEKS

MVP-1: CI Green / Test Stability (2-4h) ✅ COMPLETE

MVP-3: Cerber Doctor Command (3h) ✅ COMPLETE

MVP-4: Guardian Fast Mode <2s (3-5h) ✅ COMPLETE

MVP-2: Resilience Wiring (2-3h) ✅ COMPLETE

MVP-5: Documentation Cleanup (1-2h)

RC1 RELEASE

PHASE 2: V2.0.0 (FULL RELEASE) - AFTER RC1 FEEDBACK

V2.0.0-Full-1: Resilience Integration (Hardened from MVP-2)

V2.0.0-Full-2: Observability + Tool Configuration (8h)

V2.0.0-Full-3: Error Handling + Output Parsing (5h)

V2.0.0-Full-4: Test Coverage Hardening (5h)

V2.0.0-Full-5: State Machine / ExecutionContext (8h)

V2.0.0 Full Release

PHASE 3: V2.0.1 (PATCH RELEASE)

V2.0.1: Bug Fixes + Micro-optimizations

PHASE 4: V2.1 (FEATURE RELEASE) - 2-3 MONTHS AFTER V2.0

V2.1-1: Auto-install Missing Tools (6h)

V2.1-2: Execution Persistence (6h)

V2.1-3: SARIF Format Output (4h)

V2.1-4: Plugin System (8h)

V2.1-5: Advanced Profiles (3h)

V2.1 Release

OVERALL METRICS

KEY CHANGES FROM MVP AUDIT

EXECUTION RULES

Uh oh!

FilesExpand file tree

EXECUTION_CHECKLIST.md

Latest commit

History

EXECUTION_CHECKLIST.md

File metadata and controls

EXECUTION CHECKLIST: MVP → V2.1 DELIVERY

PHASE 1: V2.0.0-RC1 (MVP RELEASE) - 2-3 WEEKS

MVP-1: CI Green / Test Stability (2-4h) ✅ COMPLETE

MVP-3: Cerber Doctor Command (3h) ✅ COMPLETE

MVP-4: Guardian Fast Mode <2s (3-5h) ✅ COMPLETE

MVP-2: Resilience Wiring (2-3h) ✅ COMPLETE

MVP-5: Documentation Cleanup (1-2h)

RC1 RELEASE

PHASE 2: V2.0.0 (FULL RELEASE) - AFTER RC1 FEEDBACK

V2.0.0-Full-1: Resilience Integration (Hardened from MVP-2)

V2.0.0-Full-2: Observability + Tool Configuration (8h)

V2.0.0-Full-3: Error Handling + Output Parsing (5h)

V2.0.0-Full-4: Test Coverage Hardening (5h)

V2.0.0-Full-5: State Machine / ExecutionContext (8h)

V2.0.0 Full Release

PHASE 3: V2.0.1 (PATCH RELEASE)

V2.0.1: Bug Fixes + Micro-optimizations

PHASE 4: V2.1 (FEATURE RELEASE) - 2-3 MONTHS AFTER V2.0

V2.1-1: Auto-install Missing Tools (6h)

V2.1-2: Execution Persistence (6h)

V2.1-3: SARIF Format Output (4h)

V2.1-4: Plugin System (8h)

V2.1-5: Advanced Profiles (3h)

V2.1 Release

OVERALL METRICS

KEY CHANGES FROM MVP AUDIT

EXECUTION RULES