Credence — Epistemic Guard MCP server

by Lakshmi-Chakradhar-Vijayarao·io.github.Lakshmi-Chakradhar-Vijayarao/credence·v1.2.5·1 stars

No reviews yet

Epistemic guard for AI coding: tracks uncertain values and blocks writes until verified.

A88/100grade A

Write a review Install

What users say

No reviews yet

Be the first

Safety scan

A88/100

full report

Adoption

New

1 stars32 downloads/wk

Reviews

Write one

Nobody has reviewed Credence — Epistemic Guard yet.

If you have run it, two minutes of your experience saves the next person an afternoon.

Credence — Epistemic Guard tools (17)

write = sends, deletes, buys or posts

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

credence_audit
Per-session epistemic timeline — all constraints (verified and unverified) in chronological order with full certainty trajectory.
credence_autoverify
Scan text for natural-language verification signals and auto-verify matching unverified constraints — zero API calls.
credence_bandit_status
Adaptive compression threshold status (Thompson sampling bandit).
credence_constraints
List all unverified constraints for a session.
credence_diff
Compare two texts for numeric contradictions.
credence_gate
Pre-execution epistemic gate (CP4): block irreversible tool calls that embed unverified constraint values.
credence_marker_health
Marker statistics from accumulated session data.
credence_memory_recall
Load project memories into a new session at session start.
credence_memory_snapshot
Persist all unverified constraints from a session as project memory.
credence_project_status
Project-wide epistemic health dashboard.
credence_register
Register an uncertain constraint in the epistemic registry.
credence_reset
Clear all constraints for a session.
credence_scan
Generation-Time Constraint Scanner (CP3): scan model output for numeric literals that match registered unverified constraints.
credence_scan_ghosts
Scan a session for ghost constraints.
credence_self_probe
Extract domain-relevant values from generated code and register them as unverified by default — zero API calls, zero model judgment.
credence_session_summary
Brief of unverified constraints for a session.
credence_verify
Mark a registered constraint as verified with its confirmed value.

Public scan report

scanner v0.1.9 · 2026-09-27 · same rubric, same numbers if you re-run it

no findings

Code scan16 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 qualitylocal package, no credentials required12/15
Maintenancelast push 74 days ago12/15
Maintainer identityregistry namespace matches repository owner; GitHub account older than a year8/10

Overall 88/100. Components that don't apply are left out of the denominator. Any critical finding is an F.Rubric Appeal a finding JSON

How Credence — Epistemic Guard compares

Against every graded MCP server, from our own scans and uptime checks. Updated after each scan run.

Safety score: 88/100; higher than 75% of 48,255 servers
Tools: 17; the typical server has 7
Tools that change things: 0%; typical: 0%. Fewer write tools means less an agent can break
GitHub stars: 1; more than 65% of 28,688 servers
Downloads a week: 32; more than 20% of 12,191 servers with a package

What its tools let an agent do: 17 read

Install Credence — Epistemic Guard in Claude Code, Cursor or VS Code

claude mcp add credence -- uvx credence-guard

Add to Cursor

What the publisher says

From the Credence — Epistemic Guard repository's README, as published. We do not edit it. Read it on GitHub

Credence

AI doesn't remember what it wasn't sure about. Credence does.

pip install credence-guard
credence demo   # 30-second smoke test, no API key required

[mcp] adds the FastMCP server for Claude Code. Core package has zero hard dependencies.

The problem

You say: "The rate limit is probably around 50 — I haven't confirmed it yet."

Fifteen turns later, Claude writes:

RATE_LIMIT = 50   # no warning. no flag. shipped.

The API rejects every request at 2am. The real limit was 10. Claude forgot you weren't sure.

This isn't hallucination. The model reproduced exactly what it read. What it read had the qualifier stripped — by context compression, fifteen turns back.

What Credence does

Tracks uncertain values the moment you state them. Blocks writes that embed those values until you confirm them.

you say "rate limit is probably 50"
    → observer registers it (before Claude responds)
    → Claude writes: RATE_LIMIT = 50  # ⚠ CREDENCE[unverified]
    → write blocked until you confirm

Every other tool warns. Credence enforces.

What it looks like

# Claude generates this. Credence intercepts before it ships.

class StripeClient:
    API_VERSION  = "2023-10-16"  # ⚠⚠ CREDENCE[stale]: API date versions change on release — verify before shipping
    RATE_LIMIT   = 100           # ⚠  CREDENCE[unverified]: I think Stripe rate limit is around 100 req/min
    TOKEN_EXPIRY = 3600          # ⚠⚠ CREDENCE[stale]: Token/session lifetime values are set by the vendor — verify
    MAX_RETRIES  = 3
    TIMEOUT_MS   = 5000

credence: blocked Edit — 2 unverified value(s)
  → I think Stripe rate limit is around 100 req/min | TOKEN_EXPIRY = 3600
  Verify first, then retry. Use credence_constraints to see all pending.

After you confirm: "Confirmed — rate limit is 100 req/min per stripe.com/docs" → gate clears.

Setup

1. Add to .mcp.json:

{ "mcpServers": { "credence": { "command": "credence-server" } } }

2. Add to .claude/settings.json:

{
  "hooks": {
    "UserPromptSubmit": [
      { "hooks": [{ "type": "command", "command": "python3 -m credence.observer" }] }
    ],
    "PreToolUse": [
      {
        "matcher": "Write|Edit|Bash|NotebookEdit",
        "hooks": [{ "type": "command", "command": "python3 -m credence.hooks" }]
      }
    ]
  }
}

Done. No API key required.

Registry: Credence creates epistemicregistry.db in your working directory. Add .db to your .gitignore, or set CREDENCE_DB=~/.credence/registry.db to keep it global.

Session tracking: Set CREDENCESESSIONID=my-project to keep constraints stable across directory changes and terminal restarts.

Event log: The gate writes block/allow events to ~/.credence/events.jsonl (local only, never sent anywhere). Set CREDENCENOLOG=1 to disable.

Constraint cap: The registry allows up to 500 constraints per session by default. Override with CREDENCEMAXCONSTRAINTS=.

How it works

Two layers, neither requires model cooperation:

Shortened. The full README is on GitHub.

Nothing above is checked by us. What we check is on the safety report.

Credence — Epistemic Guard: common questions

Is Credence — Epistemic Guard MCP server safe?: Yes, by our scan: it is graded A (88/100). Read the Credence — Epistemic Guard safety report
How do I install Credence — Epistemic Guard?: It runs on your machine. Copy the Claude Code, Cursor, VS Code or Claude Desktop config from the install section.
Does Credence — Epistemic Guard need an API key?: Not as far as the registry entry and our scan can tell: no credentials are declared or required.
Is Credence — Epistemic Guard maintained?: The last commit was 111 days ago (2026-07-16). The latest release is v1.2.5.
How does Credence — Epistemic Guard compare with other MCP servers?: safety score 88/100 (higher than 75% of 48,255 servers); tools 17 (the typical server has 7); tools that change things 0% (typical: 0%. Fewer write tools means less an agent can break). Full comparison.