Credence — Epistemic Guard runs on your own machine — the client starts it, so there is no endpoint to ping. 79 installs a week from pypi. Last commit 1 Nov 2026.
Epistemic guard for AI coding: tracks uncertain values and blocks writes until verified.
Today is the operative word: we check Credence — Epistemic Guard every 15 minutes and re-read its code on every release. Watch it and you find out the day that stops being true.
This server runs on your own machine — install it with the package manager and the client starts it for you. Package name taken from the official registry entry.
claude mcp add credence -- uvx credence-guard
{
"mcpServers": {
"credence": {
"command": "uvx",
"args": [
"credence-guard"
]
}
}
}
[mcp_servers.credence]
command = "uvx"
args = ["credence-guard"]
{
"mcpServers": {
"credence": {
"command": "uvx",
"args": [
"credence-guard"
]
}
}
}
{
"servers": {
"credence": {
"type": "stdio",
"command": "uvx",
"args": [
"credence-guard"
]
}
}
}
Epistemic foraging for AI agents: source discovery, verified evidence, uncertainty-aware briefs
Deterministic preflight checks for credentials, fund transfers, writes, and prompt overrides.
Send AI-created PDFs for signature, track signers, and return verifiable document evidence.
Trade-verified steel grade properties, substitution verdicts, and mill-cert guidance for AI agents.
Governance runtime for compliance: verified, human-approved writes to a tamper-evident record.
Hand tasks, bugs and finished work to AI coding agents, and get back a write-up with evidence.
Find B2B leads with verified emails, write sequences, and launch and track cold email campaigns.
Evidence-chain whiteboard for AI research agents: every claim traces back to its run and code.
Answers built from our own checks of this server.