DashClaw
Fail-closed approval layer for unattended AI agents with signed audit trail.
DashClaw is the essential fail-closed rail for autonomous agent runs—if you're shipping coding agents to production, it's a must-have. Unlike observability tools like LangSmith or Langfuse that only record what happened, DashClaw intercepts before execution and forces human approval on high-risk actions. The calibration controller prevents alert fatigue, and the liveness probe ensures the governor is always awake. It's not a fully managed service—you self-host and author policies—but the 30-day
Verified 6d ago · liveness 78/100 · cite: rightaichoice.com/tools/dashclaw
- DevOps teams deploying autonomous coding agents (Claude Code, Codex) to production
- Platform teams building internal agent frameworks that need safety and auditability
- Compliance officers requiring verifiable audit trails for every AI action
- Enterprise teams running coding agents at scale where mistakes are costly
- Hobbyists building simple chatbots with minimal risk
- Users seeking a fully managed, zero-infrastructure solution (trial is limited; must self-host)
- Teams needing extensive pre-built guardrails for every scenario (requires policy authoring)
We scan live Reddit threads, YouTube comments, X posts, G2 reviews and other communities — and hand you an honest verdict in under a minute.
- Honest verdict, not marketing
- Real pros & cons from real users
- Attributed quotes with receipts
3 free scans · no card needed
Skip DashClaw if you need a fully managed, zero-infrastructure solution or you're a hobbyist with a simple chatbot—you'll spend more time configuring policies and self-hosting than you save.
The hosted tiers have monthly governed-action ceilings (sized generously, but exact numbers live in the docs) — if you exceed them, you may need to upgrade or self-host.
DashClaw's self-hosted tier is $0 forever, which undercuts every managed alternative like Apono or Torq by a wide margin — you just run it yourself. Hosted Indie ($49/mo) and Team ($199/mo) are cheaper than per-seat enterprise governance tools (e.g., Apono starts around $100/seat), and the self-hosted option fits startups and SMBs that want fail-closed safety without per-seat costs. If you need turnkey, it's pricier than pure observability tools like LangSmith's free tier.
In short
DashClaw — Fail-closed approval layer for unattended AI agents with signed audit trail. Best for DevOps teams deploying autonomous coding agents (Claude Code, Codex) to production, Platform teams building internal agent frameworks that need safety and auditability, Compliance officers requiring verifiable audit trails for every AI action. Free to start; paid plans from $49/mo.
What people actually say about DashClaw — is it worth it?
We ran a structured research pass across product reviews, community discussions, and post-purchase forum threads to surface the patterns vendors won't publish themselves. Below: the recurring strengths, the hidden costs people mention most, and the cohort that consistently regrets adopting this tool.
32 mentions across 3 sources (Hacker News, YouTube, Lemmy) · researched Aug 13, 2026.
- +Intercepts before execution, so dangerous actions can't slip through
- +Signed, replayable audit trail with Ed25519 for accountability
- +One-command install for Claude Code, Codex, and Hermes
- +Shadow mode lets you test policies without blocking anything
- +Calibration controller tunes false-block rate to your target
- −Extremely early stage with almost zero independent community feedback
- −No real-world reliability data for long unattended sessions
- −Policy tuning requires nuance to avoid false-blocks or real breaks
- −Docs and examples are mentioned but not widely shared or reviewed
- −YouTube coverage focuses on OpenClaw dashboard, not the approval layer
- • Self-hosting requires your own server, storage, and ops time
- • Potential ingestion costs for audit logs if you use cloud storage
- • Customizing policies and integrating with non-MCP agents may need developer time
Viability Score
How well maintained and how widely used is DashClaw? Built from what the vendor actually publishes (docs, changelog, tutorials, integrations, pricing), whether the site is live, and how much real users discuss it. How we calculate this
Last calculated: September 2026
How we score →Key Features
- PreToolUse hook interception for Claude Code, Codex, Hermes
- Risk decision lattice: allow, warn, allow_contained, require_approval, block
- Human-in-the-loop approval from phone, web, or inbox
- Signed, replayable audit ledger (Ed25519, JWKS)
- Calibration controller with target false-block bound
- Shadow mode for pre-enforcement testing
- Liveness probe via synthetic held action
- One-command install for Claude Code, Codex, Hermes (npx dashclaw up)
- MCP server with 17 governance tools and 3 resources
- Node.js and Python SDKs (full parity)
- REST API and CLI for custom integration
- OAuth connector for Claude Desktop (DCR + PKCE)
- Spend governance with budget tiers
- Staged workflow: allow_contained with isolated worktree
- Plan submission for long runs: one review card, per-step verdicts
About DashClaw
DashClaw is a self-hosted, MIT-licensed governance runtime that intercepts AI agent actions before they execute, freezing dangerous ones until a human approves or denies from anywhere. Built for unattended runs where native permission prompts fail because they need you at the keyboard, DashClaw protects the long, autonomous sessions of coding agents like Claude Code, Codex, and Hermes, plus chat-based assistants via OpenAI, LangChain, CrewAI, and other MCP clients. It's the difference between logging what an agent did and deciding what it's allowed to do. The core loop is intercept, decide, approve, prove. A PreToolUse hook catches a tool call, the guard engine risk-scores it against your policies into a lattice (allow, warn, allow_contained, require_approval, block), and require_approval holds the action until a human clicks approve or deny from a phone, web, or the Approvals inbox. Every decision writes a durable, replayable, signed audit row (Ed25519, JWKS). A liveness probe pushes a synthetic held action through the real hook seam to verify enforcement is still awake—stale never renders green. DashClaw's calibration controller tunes interruption frequency to a target false-block bound, running in shadow first and only tightening when active. Staged workflows let mid-band-risk file edits proceed in an isolated worktree (allow_contained), and long runs can submit their whole plan upfront for one review card with per-step verdicts. Installation is one command for Claude Code, Codex, or Hermes; an MCP server provides 17 tools and 3 resources (v5.3.0); Node and Python SDKs offer explicit control. Self-hosting is free forever—MIT licensed with no seat limits, no action ceilings, and no calling home. You own all data. Paid hosted tiers (Indie $49/mo, Team $199/mo) simply run the same stack for you, with managed infrastructure and support. A 30-day hosted trial is free, no card. Compared to observability tools like LangSmith or Langfuse that record agent actions
Behind the Verdict
DashClaw fills a gap that observability tools leave wide open. LangSmith and Langfuse show you what an agent already did; DashClaw decides what it's allowed to do before it touches anything. If you're running Claude Code or Codex unattended, that distinction is the whole ballgame. The liveness probe is the quiet killer feature. Other tools will happily report a healthy ledger while the enforcement hook silently broke hours ago. DashClaw actively tests the seam with a synthetic action and verifies it actually got blocked. That's the difference between a safety system you trust and one you hope is working. The calibration controller is why you won't rage-quit DashClaw after a week. It learns from your approve/deny stream and tunes the interruption threshold with a proven false-block bound. It runs in shadow first, then loosens as easily as it tightens—so the actions you keep approving stop asking, and one deny tightens the band immediately. This is the anti-nagging feature that keeps governance sustainable. Self-hosting is the real deal: MIT licensed, no seat limits, no action ceilings, no telemetry, no calling home. You own everything. That's rare in the AI governance space where most tools are per-seat SaaS with data going through their cloud. For compliance-heavy organizations, that's not a nice-to-have—it's the whole point. The hosted tiers are priced for what they are: managed infrastructure. Indie at $49/mo for 2 seats, Team at $199/mo for 10 seats, both with generous action ceilings and a 30-day free trial. You're not buying features—you're buying us running the stack for you. The 30-day trial is a real workspace, so you can test the flow before committing. Where it bites: DashClaw isn't turnkey. You self-host, author policies, and wire the SDK or CLI. The
Researching DashClaw? Get your full AI stack in 60 seconds.
Free, no signup — tell us your goal and get tools matched to your budget & existing stack.
Real-world workflow fit
Concrete scenarios for the personas DashClaw actually fits — and what changes day-one when you adopt it.
Setting up DashClaw to guard Claude Code deployments to production
Outcome: Within 30 minutes, you run 'npx dashclaw up', wire the PreToolUse hook, and define a policy that requires approval for any 'deploy' action with risk_score > 80. The next time Claude Code tries to deploy, the action freezes and you approve from your phone via Discord. Every decision is signed and logged.
Integrating DashClaw with an internal agent framework built on LangChain
Outcome: You use the Node.js SDK to wrap your agents' tool calls with claw.guard(). High-risk actions (e.g., deleting a database) are held for approval by the on-call team via the Approvals inbox. You enable shadow mode for a week, tune the calibration target, then switch to active enforcement. The audit ledger satisfies your
Generating audit trails for autonomous agent runs
Outcome: You connect DashClaw to your existing agents, set up agent identity pairings, and export decision history as a single file for auditors. The signed, replayable ledger proves every action was either approved or blocked, with timestamps and approver IDs. You also scan prompts for injection attempts using the built-in
Use Cases
- Intercept and block high-risk agent actions like deleting production databases
- Require human approval before deploying code to production via AI agents
- Generate audit-ready decision trails for compliance in regulated environments
- Enforce spending limits and budget tiers for agents using paid APIs
- Scan agent prompts for injection attacks and block malicious inputs
- Manage agent identities and reputations across multi-agent swarms
- Route unattended agent runs to a mobile approval inbox
- Staged file edits in an isolated worktree, promote or discard later
Models Under the Hood
as of 2026-09-01
Limitations
- DashClaw is a fail-closed approval layer that requires self-hosting the MIT-licensed runtime for full governance; the hosted trial offers a browser-based preview without installation.
- Enforcement is applied at the hook seam for Claude Code, Codex, and Hermes, and at the OpenClaw gateway, while SDKs and the MCP server honor governance cooperatively.
- Self-hosting is free forever, but hosted plans have seat limits, monthly action ceilings, and data retention windows.
- The docs indicate Node.js and Python SDK parity, with no explicit rate limits documented; performance depends on self-hosted infrastructure.
as of 2026-08-19
Verification history
We have re-verified DashClaw 6 times since . Each pass re-reads the vendor's own pages and re-checks every listed field against that evidence; passes where nothing had changed are marked as such.
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
Free to cite with attribution — this page re-verifies continuously.
12-month cost
Project the real annual outlay, including the implied monthly cost when only an annual tier is published.
Vendor list price only. Add-on usage, seat overages, and contract minimums are surfaced under Hidden costs & gotchas.
Plans compared
For each published DashClaw tier: who it actually fits, and what it adds vs. the previous tier. Cross-reference the cost calculator above for projected annual outlay.
Self-hosted
$0/mo
Ideal for
Teams that want full control and zero per-seat costs — startups, enterprises with existing infrastructure, and anyone who needs to own all data.
What this tier adds
Starting tier: free forever, MIT-licensed, includes every governance capability (guard, approvals, audit ledger, calibration) with no seat limits or action ceilings.
Hosted Indie
$49/mo
Ideal for
Solo builders or small teams who don't want to run infrastructure but need the same governance capabilities.
What this tier adds
$49/mo adds 2 seats, a monthly governed-action ceiling, 30-day retention, and email support — buys managed infrastructure, not new features.
Hosted Team
$199/mo
Ideal for
Growing teams that need more seats, higher action volumes, and priority support without managing infrastructure.
What this tier adds
$199/mo upgrades to 10 seats, a higher action ceiling, 90-day retention, org-scoped rate limits, and priority support.
Where the pricing makes sense
The company stage and team size where DashClaw's pricing actually pencils out — and where peers do it cheaper.
DashClaw's self-hosted tier is $0 forever, which undercuts every managed alternative like Apono or Torq by a wide margin — you just run it yourself. Hosted Indie ($49/mo) and Team ($199/mo) are cheaper than per-seat enterprise governance tools (e.g., Apono starts around $100/seat), and the self-hosted option fits startups and SMBs that want fail-closed safety without per-seat costs. If you need turnkey, it's pricier than pure observability tools like LangSmith's free tier.
Setup time & first value
How long it actually takes to get something useful out of DashClaw — broken out by persona, not the marketing-page minute.
For a DevOps engineer: under 30 minutes to first governed action — run 'npx dashclaw up', wire the hook, and test a guard call. For a platform team integrating SDKs: a few hours to wrap existing agents and tune policies. For a non-technical user: the hosted trial takes 5 minutes to see the Approvals inbox, but full self-hosted setup requires CLI familiarity.
Switching to or from DashClaw
How to bring data in from common predecessors and how to get it back out — written for the switcher, not the buyer.
- →From LangSmith/Langfuse (observability only): DashClaw adds active enforcement on top — set up policies and hooks; your existing logs don't transfer, but you start capturing signed decisions immediately.
- →From native permission prompts (e.g., Claude Code): DashClaw extends coverage to unattended runs; keep prompts for at-keyboard sessions, and add DashClaw for the ones you walk away from.
- ↗To Apono or Torq (managed agent governance): export your DashClaw policy set and decision history (single file) to inform your new configuration; expect per-seat pricing.
- ↗To a custom in-house guard using open-source policy engines: export your policies and audit trail; DashClaw's signed ledger gives you a starting point for your own audit format.
Integrations
Resources & Guides
Tutorials & Learning
Official links
Featured Head-to-Head Comparisons
Dashclaw vs Audioeye
AudioEye and DashClaw serve entirely different domains: AudioEye is a paid compliance platform for web accessibility, while DashClaw is an open-source governance runtime for AI agents. Choose AudioEye if you need ADA/WCAG compliance, VPAT documentation, and legal support for your website; choose DashClaw if you deploy autonomous coding agents (like Claude Code) and need guardrails, audit trails, and human-in-the-loop approval. They are not competitors.
Dashclaw vs Push Security
Choose Push Security if your priority is defending against browser-based attacks (AiTM, ClickFix, session hijacking) and controlling AI data leakage from employee browsers. Choose DashClaw if you need a governance layer for autonomous AI agents (coding agents, LLM integrations) to enforce policies, require human approval, and maintain an audit trail. They solve different problems — browser security vs. agent runtime governance — and are complementary, not competitive.
Dashclaw vs Sublime Security
If you need to govern autonomous coding agents with audit trails and human approval, choose DashClaw (open-source, self-hosted). If you need to protect your enterprise email from sophisticated BEC and phishing attacks with low false positives, choose Sublime Security (paid, integrates with M365/Google Workspace). There is no overlap in problem space.
Popular in AI Governance & Guardrails
Mindgard
Automated AI red teaming platform that continuously discovers, assesses, and defends AI systems and agents.
Poolside AI
Open-weight agentic coding models for secure on-prem enterprise AI
Olas Network
Co-own and monetize AI agents on-chain with Olas.
Frequently Asked Questions
Categories
Used DashClaw? Help shape our editorial sentiment research.