What Tablif read on Krites's site
Krites describes itself as: “Krites: verification for coding agents”
- Price model: Yearly · free tier
- Free tier: Free forever
- Stage: Beta
What it is
Calls itself “plugin”.
- Calls itself: plugin
“The Claude Code plugin: protected paths, the stop gate, signed receipts, and the verifier.”
Where it puts itself
Rules out “meters a token”.
- Says it will not: meters a token
“Krites never holds a vendor credential and never meters a token.”
What it costs
Yearly plan. A free tier, free forever. No monthly plan. Prices quoted: $240/year, $99/year.
- Monthly: no
- Yearly: yes
- Free plan: free forever
- Price: $240/year · $99/year
“Pro planned at $240/yr”
“$99 / year, locked Pro planned at $240/yr”
“The plugin is free.”
Who it sells to
Sells to developers.
- Sells to: developers
“For anyone paying for more than one agent plan.”
How it runs
A browser extension. No API. Open source. AI is the core of the product.
- Kind: extension
- API: no
- Open source: yes
- AI role: core to the product
“$ claude plugin install krites@anthropic-plugin-directory Copy”
“Krites never holds a vendor credential and never meters a token.”
What it works with
Runs inside Claude Code. Names Codex on its pages.
- Runs inside: Claude Code
- Names: Codex
“The same gate on Codex and OpenCode.”
“Works with Claude Code today.”
What it offers as proof
In beta. Claims “100”. Try it without signing up.
- Stage: beta
- Number claimed: 100
- Instant signup: yes
- Trial asks for: nothing
“Pro IN DEVELOPMENT”
“Claim a founding seat · $99”
Sources
Tablif read these pages, last on 20 Oct 2026. Everything above is what the product says about itself; Tablif does not verify every claim.
Similar products in Tablif Town
- groats: Your agent only sees tools that work.
- OpenHarnX: Your coding agent says tests pass. Prove it.
- @gent: Email a task to run your coding agent
- onsystem: Keep coding agents on your design system
- MockAgent: Mock APIs and tool calls for AI agents, no backend needed.
- The Agentic Coding Kit: Rule files that make AI coding agents ship trustworthy code
- Osade: Run all your coding agents together on real repositories
- Off the Wall: Bring past corrections back to your coding agent