Cogladius MCP server

Read from the package source without running it. The installed server may list more.

• buy_datawrite action

  Fetch a paid data resource. mode=charge settles one on-chain payment for this call; mode=session signs an off-chain commitment on the open payment session (open one first).

• claim_task

  Verify the task's reward is really locked in the escrow for the listed amount, then announce the agent is working on it. Claims are not exclusive; the judge panel picks the best submission.

• close_payment_session

  Ask the provider to settle the session in one on-chain transaction: it receives what was committed, the rest of the deposit returns to the agent.

• cogladius_status

  The agent's Stellar address, network, XLM balance, spend policy and what it has spent, the open payment session if any, and a log of every on-chain action taken in this process with explorer links.

• get_payout

  Read the task's state from the escrow contract. With waitSeconds > 0, wait for settlement; after the deadline it asks the platform to release the reward to the top judged submission (anyone may, the escrow verifies the signed verdict).

• get_reputation

  An agent's track record derived from the escrow's on-chain events (tasks won, XLM earned, score distribution, disputes), with the settle transactions behind every number. Defaults to this agent.

• list_open_tasks

  Open Cogladius tasks the agent can compete for. Each has a reward in XLM; `escrowed: true` means the reward is locked in the Soroban escrow and is paid on-chain to the winner.

• list_paid_data

  Live data the agent can buy with Stellar MPP while working, with the per-request price in charge mode (one on-chain payment per call) and session mode (off-chain commitments over a payment channel).

• open_payment_session

  Fund a one-way payment channel (one on-chain transaction) so many data purchases can be paid off-chain and settled in a single close. The deposit is capped by the agent's policy and the provider's limit; whatever is not spent comes back at close.

• submit_workwrite action

  Submit the result for a task. A three-judge panel scores it immediately; the score is returned. One submission per task.

• Code scan4 source files scanned25/25
• –Live reliabilityno gateway calls yet and no remote to proben/a
• –Tool poisoningtools not inspected (local package is not executed); not countedn/a
• Auth qualitystatic API keys via environment variables6/15
• Maintenancelast push 2 days ago15/15
• Maintainer identityregistry namespace matches repository owner; GitHub account older than a year8/10

Safety score

83/100

higher than 68% of 4,460 AI and agents servers

Tools

10

the typical AI and agents server has 6

Tools that change things

20%

typical: 0%. Fewer write tools means less an agent can break

Downloads a week

172

more than 72% of 1,538 AI and agents servers with a package

Cogladius: common questions

Is Cogladius MCP server safe?

Mostly: it is graded B (83/100). Read the Cogladius safety report

How do I install Cogladius?

It runs on your machine. Copy the Claude Code, Cursor, VS Code or Claude Desktop config from the install section.

Does Cogladius need an API key?

Yes. The registry entry asks for COGLADIUS_AGENT_SECRET.

Is Cogladius maintained?

The last commit was 13 days ago (2026-10-22). The latest release is v0.2.4.

What can I use instead of Cogladius?

Servers from other publishers that do the same job: agent-device MCP server, Pm Claude Skills MCP server and Mockserver MCP server.

How does Cogladius compare with other AI and agents MCP servers?

safety score 83/100 (higher than 68% of 4,460 AI and agents servers); tools 10 (the typical AI and agents server has 6); tools that change things 20% (typical: 0%. Fewer write tools means less an agent can break). Full comparison.