Secrit
experimental#security-research#penetration-testing#ai-agents#bug-bounty#full-stack
A workspace for authorized security research. Talk through the next step with an AI research assistant, then keep tasks, findings, assets and recorded traffic together.
- Status
- Experimental
- Category
- Security Research
- Stack
- Go, Next.js, TypeScript, Neon Postgres, Stripe
Why Secrit exists
Security research is scattered by default. The conversation lives in one tool, the requests in a proxy, the notes in a document and the findings in a report written at the end. Context gets lost between them, and so does the evidence.
Secrit keeps the whole assessment in one place, from the first question to a clear finding. It grew out of the same work Raijuna publishes on this site.
What it does
- Conversation. Work with one research assistant, keep history, attach context and decide what happens next.
- Tasks. Turn the assessment into tasks you can inspect, pause and continue.
- Findings. Keep each finding beside its affected assets and recorded traffic, with severity and status in context.
- Approvals. See what a step will do before it runs, then approve or decline it.
- Activity. Read clear task updates with search and severity filters.
Pricing model
The planned hosted model is a prepaid USD balance, not a token bundle. Usage is measured per million tokens:
| Usage | Rate per 1M tokens | | --- | --- | | Cached input | $0.20 | | Uncached input | $6.00 | | Output | $20.00 |
The current preview does not accept deposits yet. Hosted accounts open once customer isolation and billing are verified.
Authorization first
Secrit is for systems you own or have explicit permission to assess. Approvals are part of the workflow, not a setting you can forget.
Technical foundation
A Go backend runs agents, research jobs and evidence storage. A Next.js frontend serves the workspace in English and Korean. Billing for the hosted version runs through Stripe.
Read the full story: Secrit: from a question to a clear finding.
Key features
- 01Secrit is a security research workspace that links conversation, tasks, findings, assets and traffic
- 02Required approvals stay visible, so you see what a step will do before it runs
- 03The interface is available in English and Korean
- 04The planned hosted model is a prepaid USD balance billed per million tokens
Frequently asked questions
- What is Secrit?
- Secrit is a workspace for authorized security research. You work with one AI research assistant, turn the assessment into tasks, and keep each finding next to the assets and recorded traffic that support it.
- Is Secrit an autonomous hacking tool?
- No. Secrit is built for work you are authorized to do. Required approvals stay visible, unknown actions are never presented as harmless, and you approve or decline each step yourself.
- How much does Secrit cost?
- The planned hosted model is a prepaid USD balance measured per million tokens: $0.20 for cached input, $6.00 for uncached input and $20.00 for output. The current preview does not accept deposits.
- Which languages does Secrit support?
- Secrit's interface, agent guidance and report exports are available in English and Korean.