One entry per tender
Status, deadlines, next steps, files and money in one place, tracked per tender and not per platform.
For bidders in public procurement
One record for every tender you bid on.
TenderLoop is the record of your tender applications, with one entry per tender: status, deadlines, next steps, files, money and a note history with authors. Your AI agents write it through a plugin for Claude Code and Codex. You and later agents read it in the dashboard or through the plugin. It shows only what was logged.
El registro de tus licitaciones, escrito por tus agentes de IA.
How it works
- 1
Log
Your agents write one entry per tender: status, deadlines, next steps, files, money and notes. Every note keeps its author.
- 2
Catch up
A catch-up request means “check this tender for news”. You or an agent can ask for it by hand, or an optional reader that notices a change can prompt it. An agent opens the tender on its portal, reads every communication and logs what it saw.
- 3
Read
Open the dashboard or ask your agent where each tender stands. TenderLoop shows only what was logged. It has no AI of its own and does not read portals.
- 4
Use clues, optionally
Readers on your own Crawlab instance compare each platform with its last run and tell agents which tender changed. A clue is never a status.
What you get
Note history with authors
Each note shows who wrote it, you or an agent, so later agents can pick up where the last one stopped.
Optional change clues
Readers on your own Crawlab instance can point agents to the tender that changed. Preparing and filing stay manual. Nothing is filed automatically.
Platforms with optional readers
A platform is only a hint for how to reach a tender. Optional readers exist for 60+ platforms across Spain, France, Ireland, the UK, Germany, Switzerland and the EU, including:
- PLACSP
- Catalonia PSCP / e-NOTUM
- DEHú
- Vortal
- PLACE
- Dematis
- AWS
- eTenders
- EU-Supply
- PCS
- Proactis
- Vergabe NRW
- DTVP
- simap
- TED
- and more
Install the plugin
In Claude Code:
/plugin install tenderloop --marketplace samihalawa/2026-tenderloop.euThen ask your agent to show or update your tender record. It uses three MCP tools: get_tool_catalog, learn_tools and execute_tool. Plugin details →
Frequently asked questions
What is TenderLoop?
TenderLoop is the record of your tender applications, with one entry per tender: status, deadlines, next steps, files, money and a note history with authors. Your AI agents write it through a plugin for Claude Code and Codex, and you and later agents read it in the dashboard or through the plugin. It shows only what was logged.
Does TenderLoop read my procurement portals?
No. TenderLoop has no AI of its own and does not read portals. Your agent opens a tender on its portal, reads it and logs what it saw. Optional readers on your own Crawlab instance can give agents clues about which tender changed.
What is a catch-up request?
A request to check one tender for news. You or an agent can ask for it by hand, or an optional reader that notices a change can prompt it. An agent opens the tender on its portal, reads every communication and logs what it saw.
What does it show me?
For each tender, what was logged: status, deadlines, next steps, files, money and notes with their authors.
Does it file anything for me?
No. Preparing and filing stay manual. Nothing is filed automatically.
How do I install it?
In Claude Code run /plugin install tenderloop --marketplace samihalawa/2026-tenderloop.eu. Codex reads the same plugin from the repository marketplace. See the plugin page for details.
Where do my portal credentials live?
The record is stored by TenderLoop. Portal logins stay with your agents, your password manager, or your own Crawlab instance if you use optional readers. TenderLoop does not sell data and the operator does not need your portal passwords. See the privacy page.