{"slug": "show-hn-planwright-a-control-plane-for-ai-coding-agents", "title": "Show HN: PlanWright – A control plane for AI coding agents", "summary": "PlanWright launches as a control plane for AI coding agents, inverting planning and acceptance ceremonies to match agent speed. The tool synthesizes raw inputs like transcripts and Slack into structured objectives, then triages acceptance by routing only judgment-critical items to humans, with every ruling cryptographically signed for auditability.", "body_md": "The Control Plane for Agent Labor\n\n# Your coding agents got 10x faster. Your pipeline didn't.\n\nSpeed moved to the middle and left the ceremonies behind. Planwright inverts the two that are still human-paced: it turns your chaos — transcripts, decks, email, Slack — into objectives, and triages acceptance so you rule only on what needs judgment. Planning and acceptance finally move at agent speed, and every ruling is signed.\n\nClaude Desktop · MCP\n\n↳ “SOC 2 export bundle” · scheduled\n\nThe Problem\n\n## You bought the agents.\n\nWhere did the speed go?\n\nCoding used to be the bottleneck. Not anymore. Now the constraint is everything you *didn't* automate:\n\nPlanning\n\nA human hand-typing tasks at UI speed to feed machines running thousands of times faster.\n\nAcceptance\n\nA senior engineer reading pull requests one at a time — which is exactly where velocity goes to die.\n\nMake the middle free and infinitely fast, and the edges become the ceiling. That's not a tooling gap. It's Amdahl's law.\n\nThe Fix\n\n## Invert the two ceremonies coding agents can't remove.\n\n### Invert planning: chaos in, objectives out.\n\nYour planning input was never tidy — it's meeting transcripts, decks, email threads, Slack. From inside Claude Desktop, Planwright synthesizes that raw signal into structured objectives with machine-checkable acceptance criteria. The old ceremony had a human write the plan and the tool record it. Planwright flips it: the agent drafts, you approve — at the speed of describing what you want.\n\n### Invert acceptance: triage, don't drown.\n\nWhen objectives complete, Planwright picks them up and triages acceptance in Claude Cowork — combing the plain, mechanical criteria and clearing what it safely can, then routing the critical, judgment-bound items to you. The machine doesn't accept; it sorts. You stop reading every diff and start ruling only on what actually needs a human, so your scarce judgment gets concentrated instead of flooded. Every ruling is signed.\n\nThe Acceptance Gate\n\n## The machine doesn't accept — it sorts.\n\nThe state machine ensures a human rules on every judgment call before work reaches “done.” Mechanical criteria get cleared automatically; critical decisions route to you. Your sign-off is the change-approval control point your SOC 2 auditor is looking for.\n\nHuman sets the objective and schedules it for agent pickup.\n\nAgent claims the work, plans, codes, and records diffs via MCP.\n\nAgent requests acceptance. A human reviews the diff, tests, and signs off.\n\nHuman approves. The audit record is sealed and hash-chained.\n\nEvery lane transition produces a hash-chained, cryptographically signed audit record. The chain is independently verifiable on the [public trust page](/trust/silverback-cto).\n\nHow It Works\n\n## Three primitives. One throughput unlock.\n\n### Turn chaos into objectives.\n\nIn Claude Desktop, Planwright synthesizes your transcripts, decks, email, and Slack into structured objectives with machine-checkable acceptance criteria. You approve intent instead of hand-authoring tasks.\n\n### Triage acceptance, don't drown in it.\n\nWhen objectives complete, Planwright combs the mechanical criteria in Claude Cowork and clears what it safely can, routing only the judgment calls to you. The machine doesn't accept — it sorts.\n\n### Every ruling is signed.\n\nMachine-cleared throughput stays accountable: each human decision is a hash-chained record, not an ad-hoc PR read.\n\nThe Board\n\n## Objectives flow. Agents execute. Humans approve.\n\nAdd webhook retry logic\n\nMigrate auth to PKCE flow\n\nImplement ECDSA P-256 audit signing\n\nAdd RLS policies for workspaces\n\nSSE transport for MCP server\n\nDatabase schema v1\n\nHow It Works\n\n## Five steps. Zero busywork.\n\n### Write an objective\n\nDescribe the outcome you want. Not the steps. Not the implementation. A clear, atomic objective your agent can own end-to-end.\n\n### Your agent claims it via MCP\n\nClaude Code connects to your Planwright board, picks up scheduled objectives, and starts working. No copy-pasting tickets. No context-switching.\n\n### Agent plans, executes, requests acceptance\n\nThe agent decomposes the objective, writes code, runs tests, and moves it to the acceptance lane. Every step recorded in the audit chain.\n\n### You review and sign off\n\nEvery acceptance is a cryptographic signature. Not a checkbox. A real approval with your identity bound to it.\n\n### Every step hash-chained and signed\n\nImmutable audit trail from objective creation to final merge. SOC 2 evidence that writes itself.\n\nCompliance\n\n## SOC 2 auditors are raising the bar on AI-generated code.\n\nUnder the 2026 Trust Services Criteria, auditors increasingly expect immutable, tamper-evident audit trails and documented change-approval controls for AI-driven changes. Most PM tools can't produce them. Planwright ships this as a core primitive — hash-chained, ECDSA P-256-signed, from the first objective to the final merge.\n\nGet Started\n\n## One command. Every agent.\n\nNo tokens. No config files to hand-edit. Run the install command — your browser opens for GitHub login, and you're connected.\n\n```\nclaude mcp add planwright --transport http https://mcp.planwright.tools/mcp\n```\n\nZero-config. Run this command and your browser opens for GitHub login. The connection is cached for future sessions.\n\nFor CI and service accounts that can't open a browser, generate a static token in Settings → MCP Token after signing in. Audit records attribute to the token creator.\n\nNeed more detail? [Read the full Claude Code setup guide →](/docs/agents/claude-code)\n\nPricing\n\n## Start free. Scale when you're ready.\n\nEvery plan includes the MCP server, audit chain, and board. Pay for seats and retention depth.\n\n### Free\n\nFor solo developers trying agent-native planning.\n\n- Up to 3 seats\n- 1 project\n- 30-day audit retention\n- Community support\n\n[Get started free](/login)\n\n### Team\n\nFor small teams shipping with agents every day.\n\n- 4–10 seats\n- Unlimited projects\n- 1-year audit retention\n- Public trust page\n- GitHub App integration\n\n[Start free trial](/login?plan=team)\n\n### Business\n\nFor teams that need compliance out of the box.\n\n- 11–50 seats\n- Unlimited projects\n- 3-year audit retention\n- Custom domain trust page\n- Priority support\n- Advanced analytics\n\n[Start free trial](/login?plan=business)\n\n### Enterprise\n\nFor organizations with regulatory requirements.\n\n- 51+ seats\n- 7-year audit retention\n- SSO / SAML (roadmap)\n- SOC 2 attestation letter (roadmap)\n- Dedicated support\n- Custom integrations\n\n[Contact sales](https://fantastical.app/toddmerrill/silverback-cto)\n\nFAQ", "url": "https://wpnews.pro/news/show-hn-planwright-a-control-plane-for-ai-coding-agents", "canonical_source": "https://planwright.tools", "published_at": "2026-07-13 19:59:50+00:00", "updated_at": "2026-07-13 20:49:38.818977+00:00", "lang": "en", "topics": ["ai-agents", "developer-tools", "ai-tools", "ai-products", "ai-infrastructure"], "entities": ["PlanWright", "Claude Desktop", "Claude Code", "Claude Cowork", "MCP", "SOC 2"], "alternates": {"html": "https://wpnews.pro/news/show-hn-planwright-a-control-plane-for-ai-coding-agents", "markdown": "https://wpnews.pro/news/show-hn-planwright-a-control-plane-for-ai-coding-agents.md", "text": "https://wpnews.pro/news/show-hn-planwright-a-control-plane-for-ai-coding-agents.txt", "jsonld": "https://wpnews.pro/news/show-hn-planwright-a-control-plane-for-ai-coding-agents.jsonld"}}