From be83dff70a88839d1f21c1016bd4a72823337377 Mon Sep 17 00:00:00 2001 From: Claude Date: Mon, 16 Feb 2026 18:59:30 +0000 Subject: [PATCH 1/5] fix(docs): correct inaccurate comparison claims against other plugins Oh-My-ClaudeCode has partial support for compaction recovery (PreCompact hook + compaction-resilient Notepad), learning (learner skill + Notepad Wisdom System), and codebase knowledge persistence (deepinit + .omc/ state). Updated three table cells from "No" to "Partial". Broadened the "What Makes This Different" framing to acknowledge agent orchestration tools (not just spec-writing tools). Removed unverifiable "only plugin" claim about compaction recovery. https://claude.ai/code/session_01PJDpKL1xZZ4Ea9EjuvdAw4 --- README.md | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/README.md b/README.md index 31b188b..e95004d 100644 --- a/README.md +++ b/README.md @@ -50,9 +50,9 @@ Specwright closes the **entire loop** — design, plan, build, verify, ship, lea | Wiring verification (orphaned code, layer violations) | **Yes** | No | No | No | | Evidence-based PRs (criterion → code + test) | **Yes** | No | No | No | | Quality gates with findings (not just badges) | **Yes** | No | Partial | DIY | -| Compaction recovery | **Yes** | No | No | No | -| Learning system (patterns promoted across sessions) | **Yes** | No | No | No | -| Codebase knowledge persistence | **Yes** | No | No | No | +| Compaction recovery | **Yes** | No | Partial | No | +| Learning system (patterns promoted across sessions) | **Yes** | No | Partial | No | +| Codebase knowledge persistence | **Yes** | No | Partial | No | | Configurability / extensibility | Moderate | Moderate | **High** — core strength | Full control | | Lightweight / low ceremony | No — opinionated by design | **Yes** | **Yes** | Varies | @@ -60,7 +60,7 @@ Every tool in this space pushes AI-assisted development forward. Specwright's fo ## What Makes This Different -Other tools in this space focus on the **specification phase** — they help you write a plan, then hand off to the AI. The hard part isn't planning. It's everything after: does the code actually do what was asked? Is it wired up? Is it secure? Can you prove it? +Other tools in this space tend to focus on the **front half** of the loop — specification authoring, agent orchestration, or planning scaffolds — then hand off to the AI. The hard part isn't planning or delegation. It's everything after: does the code actually do what was asked? Is it wired up? Is it secure? Can you prove it? Specwright focuses on the **verification and evidence** side — the part where AI agents actually fail. @@ -74,7 +74,7 @@ Specwright focuses on the **verification and evidence** side — the part where **Codebase Health Checks** — Run `/sw-audit` periodically to find systemic issues that per-change gates miss: architecture debt, complexity growth, convention drift, accumulated workarounds. Findings persist in `AUDIT.md` with stable IDs across re-runs. Design phases surface relevant findings. The learn phase resolves them when addressed. -**Compaction Recovery** — All stateful skills support resume-from-crash. When Claude's context window compacts, Specwright reloads full state from disk. The only plugin in this space that handles context loss gracefully. +**Compaction Recovery** — All stateful skills support resume-from-crash. When Claude's context window compacts, Specwright reloads full state from disk — including workflow stage, work unit queue, and gate progress — so no manual re-orientation is needed. ## How It Works From 981b02c92cbe437a56db0e36a6cdd3b0801d8e8a Mon Sep 17 00:00:00 2001 From: Claude Date: Mon, 16 Feb 2026 19:06:25 +0000 Subject: [PATCH 2/5] docs: add Superpowers to comparison table, upgrade OMC ratings Add obra/superpowers as a new column in the comparison table. Superpowers has partial spec writing (brainstorming + plans), partial quality gates (verification-before-completion + severity-based review), partial learning (conversation memory via SQLite), high extensibility (composable SKILL.md files + community skills repo), and lightweight ceremony. Upgrade Oh-My-ClaudeCode compaction recovery, learning system, and codebase knowledge persistence from Partial to Yes per maintainer review. https://claude.ai/code/session_01PJDpKL1xZZ4Ea9EjuvdAw4 --- README.md | 24 ++++++++++++------------ 1 file changed, 12 insertions(+), 12 deletions(-) diff --git a/README.md b/README.md index e95004d..715a7d3 100644 --- a/README.md +++ b/README.md @@ -43,18 +43,18 @@ Specwright closes the **entire loop** — design, plan, build, verify, ship, lea ### How It Compares -| Capability | Specwright | [Spec Kit](https://github.com/github/spec-kit) | [Oh-My-ClaudeCode](https://github.com/Yeachan-Heo/oh-my-claudecode) | Manual workflows | -|---|---|---|---|---| -| Structured spec writing | Yes | **Yes** — core strength | Partial | DIY | -| Adversarial TDD (separate tester/executor) | **Yes** | No | No | No | -| Wiring verification (orphaned code, layer violations) | **Yes** | No | No | No | -| Evidence-based PRs (criterion → code + test) | **Yes** | No | No | No | -| Quality gates with findings (not just badges) | **Yes** | No | Partial | DIY | -| Compaction recovery | **Yes** | No | Partial | No | -| Learning system (patterns promoted across sessions) | **Yes** | No | Partial | No | -| Codebase knowledge persistence | **Yes** | No | Partial | No | -| Configurability / extensibility | Moderate | Moderate | **High** — core strength | Full control | -| Lightweight / low ceremony | No — opinionated by design | **Yes** | **Yes** | Varies | +| Capability | Specwright | [Spec Kit](https://github.com/github/spec-kit) | [Oh-My-ClaudeCode](https://github.com/Yeachan-Heo/oh-my-claudecode) | [Superpowers](https://github.com/obra/superpowers) | Manual workflows | +|---|---|---|---|---|---| +| Structured spec writing | Yes | **Yes** — core strength | Partial | Partial | DIY | +| Adversarial TDD (separate tester/executor) | **Yes** | No | No | No | No | +| Wiring verification (orphaned code, layer violations) | **Yes** | No | No | No | No | +| Evidence-based PRs (criterion → code + test) | **Yes** | No | No | No | No | +| Quality gates with findings (not just badges) | **Yes** | No | Partial | Partial | DIY | +| Compaction recovery | **Yes** | No | Yes | No | No | +| Learning system (patterns promoted across sessions) | **Yes** | No | Yes | Partial | No | +| Codebase knowledge persistence | **Yes** | No | Yes | No | No | +| Configurability / extensibility | Moderate | Moderate | **High** — core strength | **High** — core strength | Full control | +| Lightweight / low ceremony | No — opinionated by design | **Yes** | **Yes** | **Yes** | Varies | Every tool in this space pushes AI-assisted development forward. Specwright's focus is the **verification and evidence gap** — the part between "tests pass" and "it actually works." From 6fd2fd1ddfd16689d5c6ef371b69fca0ff0a582f Mon Sep 17 00:00:00 2001 From: Claude Date: Mon, 16 Feb 2026 19:12:17 +0000 Subject: [PATCH 3/5] docs: upgrade comparison ratings where feature evidence exists MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Give credit where evidence exists — if a tool has the feature, mark Yes: - OMC: structured specs (PRD + ralplan), TDD (tdd skill + critic), quality gates (4-phase verifier with 7 checks) - Superpowers: structured specs (brainstorming + writing-plans), TDD (mandatory red-green-refactor + two-stage review), quality gates (verification-before-completion), learning (SQLite memory) - Spec Kit: quality gates (/speckit.checklist) Only "No" remains where there is genuinely no evidence of the feature. https://claude.ai/code/session_01PJDpKL1xZZ4Ea9EjuvdAw4 --- README.md | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/README.md b/README.md index 715a7d3..0e6a2f8 100644 --- a/README.md +++ b/README.md @@ -45,13 +45,13 @@ Specwright closes the **entire loop** — design, plan, build, verify, ship, lea | Capability | Specwright | [Spec Kit](https://github.com/github/spec-kit) | [Oh-My-ClaudeCode](https://github.com/Yeachan-Heo/oh-my-claudecode) | [Superpowers](https://github.com/obra/superpowers) | Manual workflows | |---|---|---|---|---|---| -| Structured spec writing | Yes | **Yes** — core strength | Partial | Partial | DIY | -| Adversarial TDD (separate tester/executor) | **Yes** | No | No | No | No | +| Structured spec writing | Yes | **Yes** — core strength | Yes | Yes | DIY | +| Adversarial TDD (separate tester/executor) | **Yes** | No | Yes | Yes | No | | Wiring verification (orphaned code, layer violations) | **Yes** | No | No | No | No | | Evidence-based PRs (criterion → code + test) | **Yes** | No | No | No | No | -| Quality gates with findings (not just badges) | **Yes** | No | Partial | Partial | DIY | +| Quality gates with findings (not just badges) | **Yes** | Yes | Yes | Yes | DIY | | Compaction recovery | **Yes** | No | Yes | No | No | -| Learning system (patterns promoted across sessions) | **Yes** | No | Yes | Partial | No | +| Learning system (patterns promoted across sessions) | **Yes** | No | Yes | Yes | No | | Codebase knowledge persistence | **Yes** | No | Yes | No | No | | Configurability / extensibility | Moderate | Moderate | **High** — core strength | **High** — core strength | Full control | | Lightweight / low ceremony | No — opinionated by design | **Yes** | **Yes** | **Yes** | Varies | From 680c5cc63e549c908b5f460ba06a1d84a5104cb0 Mon Sep 17 00:00:00 2001 From: Claude Date: Mon, 16 Feb 2026 19:14:35 +0000 Subject: [PATCH 4/5] docs: downgrade Spec Kit and Superpowers quality gates to Partial Spec Kit has checklists but no structured findings or default-FAIL semantics. Superpowers has verification-before-completion but no categorized findings documents. https://claude.ai/code/session_01PJDpKL1xZZ4Ea9EjuvdAw4 --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index 0e6a2f8..edd9406 100644 --- a/README.md +++ b/README.md @@ -49,7 +49,7 @@ Specwright closes the **entire loop** — design, plan, build, verify, ship, lea | Adversarial TDD (separate tester/executor) | **Yes** | No | Yes | Yes | No | | Wiring verification (orphaned code, layer violations) | **Yes** | No | No | No | No | | Evidence-based PRs (criterion → code + test) | **Yes** | No | No | No | No | -| Quality gates with findings (not just badges) | **Yes** | Yes | Yes | Yes | DIY | +| Quality gates with findings (not just badges) | **Yes** | Partial | Yes | Partial | DIY | | Compaction recovery | **Yes** | No | Yes | No | No | | Learning system (patterns promoted across sessions) | **Yes** | No | Yes | Yes | No | | Codebase knowledge persistence | **Yes** | No | Yes | No | No | From 76da50c16e22265e018a414979ba593dcfb1f994 Mon Sep 17 00:00:00 2001 From: Claude Date: Mon, 16 Feb 2026 19:17:29 +0000 Subject: [PATCH 5/5] docs: remove configurability and ceremony rows from comparison table https://claude.ai/code/session_01PJDpKL1xZZ4Ea9EjuvdAw4 --- README.md | 2 -- 1 file changed, 2 deletions(-) diff --git a/README.md b/README.md index edd9406..d61ac12 100644 --- a/README.md +++ b/README.md @@ -53,8 +53,6 @@ Specwright closes the **entire loop** — design, plan, build, verify, ship, lea | Compaction recovery | **Yes** | No | Yes | No | No | | Learning system (patterns promoted across sessions) | **Yes** | No | Yes | Yes | No | | Codebase knowledge persistence | **Yes** | No | Yes | No | No | -| Configurability / extensibility | Moderate | Moderate | **High** — core strength | **High** — core strength | Full control | -| Lightweight / low ceremony | No — opinionated by design | **Yes** | **Yes** | **Yes** | Varies | Every tool in this space pushes AI-assisted development forward. Specwright's focus is the **verification and evidence gap** — the part between "tests pass" and "it actually works."