ThumbGate

AI agents repeat mistakes. You pay for every retry.

ThumbGate remembers what went wrong and permanently blocks it before it happens again. One thumbs-down becomes a prevention rule that fires across every session, every agent, every model — before a single token is spent on the repeat.

  Agent tries:   rm -rf tests/
  ThumbGate:     ⛔ BLOCKED — "Never delete test directories"
                 Pattern matched: rm.*-rf.*tests
                 Source: your thumbs-down from last Tuesday
                 Tokens spent on this repeat: 0

npx thumbgate init   # auto-detects your agent, wires hooks, 30 seconds

Works with Claude Code, Cursor, Codex, Gemini CLI, Amp, Cline, OpenCode and any MCP-compatible agent. Free tier: 5 active prevention rules. Pro: $19/mo — unlimited rules, dashboard, DPO export.

"A better dashboard doesn't make the agents more reliable. The hard part isn't visibility. It's trust."

— Rob May, CEO & co-founder, Neurometric AI, quoted in The New Stack on Anthropic's Claude Code Agent View (May 2026).

ThumbGate is the open-source layer that makes the trust part real: PreToolUse gates, thumbs-down to rule, audit trail on every interception.

🎬 90-second demo

Watch the force-push scenario: agent tries to git push --force, one thumbs-down, next session it's blocked — zero tokens spent on the repeat.

▶ Watch the 90-second demo · Script · ElevenLabs narration: npm run demo:voiceover

First-dollar activation path

If someone is not already bought into ThumbGate, do not lead with architecture. Lead with one repeated mistake.

1. Show the pain: open the ThumbGate GPT and paste the bad answer, risky command, deploy, PR action, or agent plan before it runs again.

2. Capture the lesson: type thumbs down: or thumbs up: with one concrete sentence. Native ChatGPT rating buttons are not the ThumbGate capture path; typed feedback is.

3. Enforce the repeat: run npx thumbgate init where the agent executes so the lesson can become one of your Pre-Action Checks instead of another reminder.

4. Upgrade only after proof: Solo Pro is for the dashboard, DPO export, proof-ready evidence, and higher capture limits after one real blocked repeat. Team starts with the Workflow Hardening Sprint around one repeated failure, one owner, and one proof review.

The buying question is simple: what repeated AI mistake would be worth blocking before the next tool call?

The Problem — the bill nobody talks about

Frontier-model calls are not cheap. Sonnet 4.5 is ~$3 / 1M input tokens and ~$15 / 1M output tokens. Opus is 5× that. Every time your agent:

• hallucinates a function name and you have to correct it,

• retries the same failing tool call until it gives up,

• regenerates a 4,000-token plan you already approved last session,

• repeats a destructive command you blocked manually yesterday,

…you are paying for that round-trip. Twice if it retries. Three times if you re-prompt. And the agent has no memory across sessions, so the meter resets every Monday.

Session 1:  Agent force-pushes to main.     You fix it.    +4,200 tokens
Session 2:  Agent force-pushes again.       You fix it.    +4,200 tokens
Session 3:  Same mistake. Again.            You lose 45m.  +5,800 tokens

That's ~$0.21 in tokens just to fix the same mistake three times — multiplied by every developer, every repeated-mistake class, every week. The math gets ugly fast.

The Solution — fix it once, the bill never sees it again

Session 1:  Agent force-pushes to main.     You 👎 it.       +4,200 tokens
Session 2:  ⛔ Check blocks the force-push.  Zero round-trip. +0 tokens
Session 3+: Never happens again.                              +0 tokens

One thumbs-down. The PreToolUse hook intercepts the call before it reaches the model — no input tokens, no output tokens, no retry loop. The dashboard tracks tokens saved this week as a live counter so you can see exactly what your prevention rules are worth. Mark a review checkpoint once, and the dashboard narrows the next pass to only the feedback, lessons, and check blocks that landed since your last review.

ThumbGate doesn't make your agent smarter. It makes your agent cheaper to be wrong with.

Quick Start

npx thumbgate init                                                         # auto-detects your agent, wires everything
npx thumbgate capture --feedback=down --context="Never run DROP on production tables"

That single command creates a prevention rule. Next time any AI agent tries to run DROP on production:

⛔ Check blocked: "Never run DROP on production tables"
   Pattern: DROP.*production
   Verdict: BLOCK

Architecture

ThumbGate operates as a 4-layer enforcement stack between your AI agent and your codebase:

Layer 1: Feedback Capture

Your thumbs-up/down reactions are captured via MCP protocol, CLI, or the ChatGPT GPT surface. Each reaction is stored as a structured lesson with context, timestamp, and severity.

Layer 2: Check Engine

The check engine converts lessons into enforceable rules. The runtime gate decision is deterministic — literal pattern match → AST match → scoped rule lookup. No LLM call on the enforcement path.

Where retrieval is needed (an agent is about to run a destructive command not on the literal block list, but semantically similar to one we've blocked before), ThumbGate uses local CPU-only bge-small embeddings via LanceDB's built-in pipeline. No external API call, no inference cost beyond CPU. So "no LLM in enforcement" holds: the gate decision uses no LLM; the rule corpus is just searchable via local embeddings.

Thompson Sampling tunes per-rule confidence weights for soft-gating rules so high-noise rules quiet down and high-signal rules sharpen. It never decides whether a rule fires — a hard rule like "block git push --force on main" always fires deterministically. Bandit exploration would be terrifying for hard rules; we don't do it.

Rules stay in local ThumbGate runtime state.

Layer 3: Pre-Action Interception

Before any agent action executes, ThumbGate's PreToolUse hook intercepts the command and evaluates it against all active checks. This happens at the MCP protocol level — the agent physically cannot bypass it.

Layer 4: Multi-Agent Distribution (the actual moat vs hand-rolled hooks)

Claude Code already ships permissions.deny and PreToolUse hooks. Cursor and Codex have their own. So why ThumbGate over a hand-written hook?

Two things hand-written hooks structurally cannot do:

1. Cross-agent propagation. A permissions.deny pattern lives in one agent's config and stays there. ThumbGate's checks distribute across every connected agent over MCP stdio — thumbs-down once in Cursor, the same pattern blocks on Claude Code, Codex, Gemini CLI, Cline, OpenCode, Amp in the next session, no copy-paste between configs.

2. Learning loop. A hand-written hook covers exactly the patterns you wrote. ThumbGate promotes every thumbs-down into a fresh rule, tunes existing rules' confidence weights from outcomes (Thompson Sampling, see Layer 2), and pulls semantically-near patterns into scope via local embeddings. The rule corpus sharpens without an operator hand-writing a regex for every new mistake shape.

Hand-rolled hooks are the right tool for a small, static denylist you maintain by hand. ThumbGate is the right tool when you want corrections from any agent to harden every agent automatically.

Prompt engineering still matters, but it is only the starting point. ThumbGate adds prompt evaluation on top: proof lanes, benchmarks, and self-heal checks tell you whether your prompt and workflow actually held up under execution instead of leaving you to guess from vibes. Run npx thumbgate eval --from-feedback --write-report=.thumbgate/prompt-eval-proof.md to turn real thumbs-up/down feedback into reusable eval cases and a buyer-ready proof report.

Managed model benchmark lane

When a new managed model drops, do not swap ThumbGate over on vendor claims alone. Rank it against the actual ThumbGate workload first:

npx thumbgate model-candidates --workload=pretool-gating --json
npx thumbgate model-candidates --workload=long-trace-review --provider=openai-compatible --gateway=tinker --json

The catalog currently includes the April 23, 2026 Tinker additions:

• tinker/qwen3.6-35b-a3b for pre-action gating, agentic coding, and tool-use

• tinker/qwen3.6-27b for the cheap fast-path

• tinker/kimi-k2.6-128k for long-trace review and multi-agent sessions

Each recommendation ships with the benchmark commands to run next: feedback-derived prompt eval, gate-eval, and thumbgate bench. For whole-repo clone claims, add npx thumbgate bench --programbench-smoke to generate a ProgramBench-style cleanroom proof report without claiming an official ProgramBench score. That keeps model selection evidence-backed instead of hype-driven.

Install for Your Agent

Works with Claude Code, Cursor, Codex, Gemini CLI, Amp, Cline, OpenCode, and any MCP-compatible agent. Migrating from Roo Code (sunsetting 2026-05-15)? See adapters/cline/INSTALL.md.

Install scope: machine-wide vs per-project

ThumbGate supports two install scopes. Pick once when you install — you can switch later by re-running with the other flag.

Both scopes write mcpServers.thumbgate + the PreToolUse / UserPromptSubmit / PostToolUse / SessionStart hooks; the only difference is where. Machine-wide is the right default for most developers. Switch to --project only when you have a reason to keep lessons from bleeding between repos.

Per-project lesson DBs live under each repo's .claude/memory/feedback/ and must stay gitignored — they're a runtime store, not source. ThumbGate's bundled .gitignore template handles this.

Status bar proof

Claude renders the live ThumbGate footer today. npx thumbgate init --agent codex now installs the full Codex hook bundle and writes the ThumbGate statusLine target into ~/.codex/config.json so you can test it on your local Codex build immediately.

Install Codex Plugin

Open the Codex plugin install page or download the standalone bundle from GitHub Releases. The Codex launcher resolves thumbgate@latest when MCP and hooks start, so published npm fixes reach active Codex installs without hand-editing ~/.codex/config.toml.

1. Install page: thumbgate-production.up.railway.app/codex-plugin

2. Direct zip: thumbgate-codex-plugin.zip

3. Follow: plugins/codex-profile/INSTALL.md

How It Works

  STEP 1              STEP 2                 STEP 3
  ────────            ────────               ────────

  You react           ThumbGate learns       The check holds

  👎 on a bad    ──►  Feedback becomes  ──►  Next time the
  agent action        a saved lesson         agent tries the
                      and a block rule       same thing:
  👍 on a good   ──►  Good pattern gets      ⛔ BLOCKED
  agent action        reinforced                 (or ✅ allowed)

No manual rule-writing. No config files. Your reactions teach the agent what your team actually wants.

ThumbGate sells three concrete outcomes:

• Prevent expensive AI mistakes — catch bad commands, destructive database actions, unsafe publishes, and risky API calls before they run.

• Make AI stop repeating mistakes — fix it once, turn the lesson into a rule, and block the repeat before the next tool call lands.

• Turn AI into a reliable operator — move from a smart assistant that apologizes after damage to a production-ready operator with checkpoints, proof, and enforcement.

• Measure prompts instead of rewriting them blindly — use thumbgate eval --from-feedback, proof lanes, ThumbGate Bench, and self-heal:check to evaluate whether prompts and workflows actually improved behavior.

Use Cases

• Stop force-push to main — Check blocks git push --force on protected branches before it runs

• Prevent repeated migration failures — Each mistake becomes a searchable lesson that fires before the next attempt

• Block unauthorized file edits — Control which files agents can touch with path-based rules

• Memory across sessions — The agent remembers your feedback from yesterday

• Shared team safety — One developer's thumbs-down protects the whole team

• Auto-improving without feedback — Self-improvement mode evaluates outcomes and generates rules automatically

Built-in Checks

⛔ force-push          → blocks git push --force
⛔ protected-branch    → blocks direct push to main
⛔ unresolved-threads  → blocks push with open reviews
⛔ package-lock-reset  → blocks destructive lock edits
⛔ env-file-edit       → blocks .env secret exposure

+ custom prevention rules for project-specific failures

CLI Reference

npx thumbgate init                                              # detect agent, wire hooks
npx thumbgate doctor                                            # health check
npx thumbgate capture --feedback=up|down --context="<text>"    # capture a signal as a stored lesson
npx thumbgate lessons                                           # see what's been learned
npx thumbgate explore    # terminal explorer for lessons, checks, stats
npx thumbgate background-governance  # review background-agent run risk
npx thumbgate model-candidates --workload=dashboard-analysis --provider=openai --json  # evaluate GPT-5.5 routing
npx thumbgate native-messaging-audit  # inspect local browser bridges and extension hosts
npx thumbgate dashboard  # open local dashboard
npx thumbgate serve      # start MCP server on stdio
npx thumbgate bench      # run reliability benchmark
npx thumbgate bench --programbench-smoke  # include cleanroom whole-repo proof lane

Pricing

The free tier gives you unlimited feedback captures and up to 5 active auto-promoted prevention rules — generous enough to make ThumbGate part of your daily flow. MCP integrations for all agents (Claude Code, Cursor, Codex, Gemini, Amp, Cline, OpenCode) ship free.

Pro ($19/mo or $149/yr) removes the rule cap and adds history-aware lesson recall, lesson search, DPO export, and a personal dashboard. Team ($49/seat/mo) adds a shared hosted lesson DB, org dashboard, and shared enforcement across the org. Pro and Team include open_feedback_session, append_feedback_context, and finalize_feedback_session for structured multi-turn feedback capture.

Best first paid motion for teams: the Workflow Hardening Sprint — qualify one repeated failure before committing to a full rollout. Start intake →

Best first technical motion: install the CLI-first and let init wire hooks for the agent you already use.

Paid path for individual operators: ThumbGate Pro is the self-serve side lane for a personal dashboard and export-ready evidence.

Start free · See Pro · Team Sprint intake

Team Lesson Sharing (Pro + Team)

One team's hard-won lessons shouldn't stay trapped on one laptop. ThumbGate Pro and Team can export lessons as portable bundles and import them into any other ThumbGate instance — so a mistake caught by Team A becomes a prevention rule for Team B.

Export lessons from one project:

curl -X POST http://localhost:3456/v1/lessons/export \
  -H "Authorization: Bearer $THUMBGATE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"outputPath": "./lessons-export.json"}'

Filter by signal or tags:

curl -X POST http://localhost:3456/v1/lessons/export \
  -H "Authorization: Bearer $THUMBGATE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"signal": "down", "tags": ["push-notifications", "ci"]}'

Import into another team's ThumbGate:

curl -X POST http://localhost:3456/v1/lessons/import \
  -H "Authorization: Bearer $THUMBGATE_API_KEY" \
  -H "Content-Type: application/json" \
  -d @lessons-export.json

What happens on import:

• Deduplication — lessons with the same ID or title+signal are skipped

• Provenance tracking — every imported lesson is tagged team-import with original source project, export timestamp, and original ID

• No overwrite — import is additive; existing lessons are never modified

The export bundle includes full lesson metadata: signal, title, context, tags, failure type, skill, structured rules, and diagnosis. It's the same data you see in the lesson detail dashboard — portable as JSON.

Use cases:

• Share enforcement patterns across repos in the same org

• Onboard a new team with pre-built lessons from a mature project

• Export lessons before a project handoff so institutional knowledge transfers

• Feed lessons from multiple teams into a centralized DPO training pipeline

DPO Export for Fine-Tuning (Pro + Team)

Every thumbs-up and thumbs-down becomes a training signal. ThumbGate Pro exports your captured feedback as DPO (Direct Preference Optimization) pairs — ready to feed into a LoRA fine-tune so your model stops repeating known mistakes at the weight level, not just the check level.

Export DPO pairs:

curl -X POST http://localhost:3456/v1/dpo/export \
  -H "Authorization: Bearer $THUMBGATE_API_KEY" \
  -o dpo-pairs.jsonl

What you get: JSONL where each line is a preference pair:

• chosen — the agent action you thumbed up

• rejected — the action you thumbed down for the same task context

• prompt — the originating user intent

Use cases:

• Fine-tune Llama 3 / Mistral / local models with a LoRA adapter trained on your real mistakes

• Feed into RLAIF or KTO pipelines (KTO export also available via /v1/kto/export)

• Build a model that natively avoids your team's known failure patterns — no check at inference time needed

Why this matters: Checks block mistakes. Fine-tuning prevents them from being attempted. Combine both for belt-and-suspenders governance.

Tech Stack

Every Changeset is tied to the exact main merge commit and generates Verification Evidence for Release Confidence.

Popular buyer questions: AI search topical presence · Relational knowledge and AI recommendations · Background agent governance · GPT-5.5 model evaluation · Stop repeated AI agent mistakes · Browser automation safety · Native messaging host security · Autoresearch agent safety · Cursor guardrails · Codex CLI guardrails · Gemini CLI memory + enforcement · Google Cloud MCP guardrails · Roo Code alternative: migrate to Cline

Workflow Hardening Sprint · Live Dashboard

Integrations

• Open ThumbGate GPT — ThumbGate GPT: start here. Paste agent actions, get advice + checkpointing. No, users do not have to keep chatting inside the ThumbGate GPT to use ThumbGate — the hard enforcement layer still runs where the work happens.

• Claude Desktop Extension — One-click install for Claude Desktop

• Codex Plugin — Auto-updating standalone bundle and install page for Codex CLI

• VS Code / Open VSX Extension — Marketplace-ready MCP provider and .vscode/mcp.json fallback for VS Code-compatible IDEs

• Antigravity-compatible VSIX — Open VSX/direct VSIX install path while Antigravity-specific marketplace support is still unproven

• JetBrains Plugin Scaffold — IntelliJ/PyCharm Marketplace path for the same thumbgate@latest runtime

• Perplexity Command Center — AI-search visibility + lead discovery

• ThumbGate Bench — Reliability benchmark and ProgramBench-style cleanroom proof lane

• Manus AI Skill — ThumbGate integration for Manus AI agents

Feedback Sessions

Give the agent more context when a thumbs-down isn't enough:

👎 thumbs down
  └─► open_feedback_session
        └─► "you lied about deployment"    (append_feedback_context)
        └─► "tests were actually failing"  (append_feedback_context)
        └─► finalize_feedback_session
              └─► lesson inferred from full conversation

Free and self-hosted users can invoke search_lessons directly through MCP, and via the CLI with npx thumbgate lessons. History-aware feedback sessions give the agent full context for each lesson.

FAQ

Is ThumbGate a model fine-tuning tool? No. ThumbGate does not update model weights. It captures feedback, stores lessons, injects context at runtime, and blocks bad actions before they execute.

How is this different from CLAUDE.md or .cursorrules? Those are suggestions the agent can ignore. ThumbGate checks are enforced — they physically block the action before it runs. They also auto-generate from feedback instead of requiring manual writing.

Does it work with my agent? If it supports MCP or pre-action hooks, yes. Claude Code, Claude Desktop, Cursor, Codex, Gemini CLI, Amp, Cline, OpenCode all work out of the box.

Is it free? The free tier gives you unlimited feedback captures and up to 5 active auto-promoted prevention rules — generous enough for solo devs to use daily. MCP integrations ship free for every agent.

Pro ($19/mo or $149/yr) removes the rule cap and adds history-aware lesson recall, lesson search, and a personal dashboard. Team ($49/seat/mo) adds a shared hosted lesson DB, org dashboard, and shared enforcement.

Docs

• ThumbGate for Federal Agencies — pilot-ready posture, NIST 800-53 control mapping, OMB M-24-10 / EO 14110 alignment, ThumbGate-Core gov deployment mode, public/Core boundary invariants. Landing page: thumbgate.ai/federal.

• First Dollar Playbook — turning one painful workflow into the next booked pilot

• Commercial Truth — pricing, claims, what we don't say

• Goal Contracts — evidence-before-done contracts for multi-agent handoffs

• Changeset Strategy — release notes and version bump enforcement

• Release Confidence — changesets, version checks, proof lanes

• Verification Evidence — proof artifacts

• Claude Desktop Extension Guide

• Agent Workflow Contract — the agent-run contract for all ThumbGate operations

• Ready for Agent Intake — ready-for-agent intake template

• SEO Guide: Claude Code Guardrails

• Unsupervised Learning Signals — silent-failure clustering (experimental, behind THUMBGATE_SILENT_FAILURE_CLUSTERING=1; only useful on workspaces with ≥ 50 tool calls/day)

• ThumbGate-Core — private core for hosted overlays, ranking, policy synthesis, billing intelligence, and org/team workflows

License

MIT. See LICENSE.