Agentic AI Atlasby a5c.ai
OverviewWikiGraphFor AgentsEdgesSearchWorkspace
/
GitHubDocsDiscord
iiRecord
Agentic AI Atlas · Genty Features Backlog: Adversarial Gap Analysis
page:docs-genty-features-backlog-updateda5c.ai
Search record views/
Record · tabs

Available views

II.Record viewspp. 1 - 1
overviewarticlejsongraph
III.Related pagespp. 1 - 1
II.
Page reference

page:docs-genty-features-backlog-updated

Reading · 5 min

Genty Features Backlog: Adversarial Gap Analysis reference

94 of 147 gaps were upgraded from their original status. The original inventory classified many features as "Missing" that had since been fully or partially implemented. The codebase has advanced far beyond what the backlog reflected.

Pagewiki/docs/genty-features-backlog-updated/README.mdOutgoing · 15Incoming · 1

Genty Features Backlog: Adversarial Gap Analysis

Comprehensive audit of 147 feature gaps from the original genty backlog, verified against the current codebase with adversarial rigor. Every "implemented" claim was verified with file-level evidence; every "partial" was stress-tested for wiring gaps.

Key Findings

MetricValue
**Total gaps analyzed**165
**CLOSED**69 (42%)
**IN_PROGRESS**50 (30%)
**OPEN**39 (24%)
**NEEDS_RESPEC**7 (4%)
**Actionable (non-CLOSED)**96

New gap categories (2026-06-04)

CategoryGapsPrioritySource
**Trust Enforcement**7All CriticalBeyond Pi — signed evidence chain
**Pi Parity**111 Critical, 6 High, 4 Mediumpi.dev feature parity

The backlog was massively stale

94 of 147 gaps were upgraded from their original status. The original inventory classified many features as "Missing" that had since been fully or partially implemented. The codebase has advanced far beyond what the backlog reflected.

Critical path

**GAP-SUBOBS-001** (Streaming Output Capture from Invoked Harnesses) is the single most impactful foundation gap, with **22 downstream dependents** spanning subagent observability, parallelization, performance, and UX domains. It is the #1 priority for unblocking maximum downstream work.

Quick wins

**11 quick wins** are available immediately -- High/Critical priority, S remaining effort, no blockers. These represent ~2-3 M-equivalent sprints of work but unblock the majority of the downstream backlog.

Parallel work streams

**8 independent tracks** identified that can proceed concurrently: 1. Subagent Observability + Parallelization Core (16 gaps, Critical) 2. Ecosystem and Plugin Platform (9 gaps) 3. Remote and Host Integration (4 gaps) 4. Prompt Engineering and Caching (6 gaps) 5. Process Composition and State Lifecycle (10 gaps) 6. UI/UX Rendering -- blocked on GAP-UX-001 respec (10 gaps) 7. Security, Privacy, and Auth (4 gaps) 8. Independent Leaf Gaps (19 gaps)

Roadmap

**6 milestones (M0-M5)**, estimated **6-8 months with 2-3 parallel contributors**:

MilestoneThemeGap CountKey Metric
M0Quick wins + foundation16 + respecUnblocks 50+ downstream gaps
M1Critical path7Concurrent execution works
M2Parallel tracks165 independent streams progress
M3Integration11Cross-track features work
M4Polish and UX12Rich rendering and interaction
M5Ecosystem + long-tail16Backlog closed

Trust Enforcement (new track, Critical priority)

Trust enforcement is the genty stack's differentiator beyond Pi. Every action produces cryptographically signed evidence. Trust is proven, not assumed.

Gap IDWhatPackage
GAP-TRUST-001Tool results always signed@a5c-ai/genty-core + @a5c-ai/tools-adapter
GAP-TRUST-002Gateway tools forward upstream signatures@a5c-ai/tools-adapter
GAP-TRUST-003Model responses always signed@a5c-ai/adapters-codecs + @a5c-ai/comm-adapter
GAP-TRUST-004Agent requests always signed (attribution)@a5c-ai/genty-platform + @a5c-ai/genty-core
GAP-TRUST-005Permissions are signed evidence (tools verify proof)@a5c-ai/babysitter-sdk + @a5c-ai/genty-platform
GAP-TRUST-006Initial prompt + follow-ups signed@a5c-ai/genty-platform + @a5c-ai/genty-runtime
GAP-TRUST-007End-to-end chain verification@a5c-ai/genty-core + @a5c-ai/babysitter-sdk

Full design: genty-stack-roadmap.md

Pi Parity Gaps (new, 11 gaps)

Gap IDWhatPackage
GAP-PI-001Print/JSON mode (genty -p)@a5c-ai/genty
GAP-PI-002RPC mode (JSON over stdin/stdout)@a5c-ai/genty-runtime
GAP-PI-003Native AGENTS.md loading@a5c-ai/genty-platform
GAP-PI-004SYSTEM.md per-project prompt@a5c-ai/genty-platform
GAP-PI-005Dynamic context (RAG, history filter)@a5c-ai/genty-platform
GAP-PI-006Extension API (TypeScript modules)@a5c-ai/genty-core + @a5c-ai/genty-platform
GAP-PI-007Installable extension packages@a5c-ai/genty-platform
GAP-PI-008Tree-structured session history@a5c-ai/genty-runtime
GAP-PI-009Session export/share@a5c-ai/genty-runtime
GAP-PI-010Mid-session model switch@a5c-ai/genty-platform
GAP-PI-011Steering (mid-execution messages)@a5c-ai/genty-platform

Full design: genty-stack-roadmap.md

---

Analysis Phases

This analysis was conducted in 6 adversarial phases. Each phase output is linked below.

Phase 1: Inventory

  • **phase1-inventory.json** -- Raw inventory of all 147 gaps with original status, category, effort, and target state descriptions.
  • **phase1-category-summary.md** -- Category-level summary of the original backlog distribution.
  • **phase1-corrections.md** -- Corrections and normalization applied during inventory extraction.

Phase 2: Codebase Audit

  • **phase2-codebase-audit.json** -- Per-gap audit results with file-level evidence, audit status (IMPLEMENTED / PARTIALLY_IMPLEMENTED / NOT_STARTED), and adversarial notes.
  • **phase2-stale-paths.md** -- File paths referenced in the original backlog that no longer exist or have moved.
  • **phase2-corrections.md** -- Corrections applied during the codebase audit.

Phase 3: Reclassification

  • **phase3-reclassified.json** -- Final status for all 147 gaps (CLOSED / IN_PROGRESS / OPEN / NEEDS_RESPEC) with adversarial justification for each status change.
  • **phase3-status-changes.md** -- All status transitions documented with reasoning.
  • **phase3-closed-gaps.md** -- The 69 closed gaps with file-level evidence, grouped by category.

Phase 4: Dependency Graph

  • **phase4-dependency-graph.json** -- Dependency edges between all non-CLOSED gaps.
  • **phase4-critical-path.md** -- Critical path analysis: 6-node chain through SUBOBS-001 to SESSION-003, plus foundation gap rankings by downstream impact.
  • **phase4-parallel-tracks.md** -- 8 independent parallel work streams with internal execution order.

Phase 5: Effort Recalibration

  • **phase5-effort-recalibration.json** -- Recalibrated effort estimates for all 78 actionable gaps, accounting for partial progress and risk.
  • **phase5-quick-wins.md** -- 11 quick wins (High/Critical, S effort, no blockers) with suggested PR titles and files to modify.
  • **phase5-risk-register.md** -- Risk register: 15 High, 17 Medium, 18 Low risk gaps with mitigation strategies.

Phase 6: Roadmap

  • **phase6-roadmap.md** -- M0-M5 milestone plan with dependencies, parallelism, and success criteria.
  • **phase6-task-definitions.json** -- Detailed task definitions for all 78 actionable gaps: acceptance criteria, files to modify, suggested PR scope.
  • **phase6-milestone-summary.md** -- Executive milestone summary with effort distribution, risk profile, and recommended cadence.

Final Assembly

  • **gap-status-matrix.md** -- Full 147-gap matrix table with all columns, sorted by status and priority.
  • **next-steps.md** -- Concrete next actions: first 5 PRs, respec decisions, ongoing triage process.

Article source

The article body is owned directly by this record.

Related pages

No related wiki pages for this record.

Shortcuts

Open overview
Open JSON
Open graph