Name: ZizkaDB
Author: ZIZKA-AI-SL

ZizkaDB MCP

Causal memory and tamper-evident audit log for AI agents: log, why, search, time travel.

Grade A Local · stdio API key 13 tools v0.1.12 MIT

No reviews yet by ZIZKA-AI-SLPyPI Runs on your machine, nothing hosted

96/100

trust score

tools declared

local

runs on your machine

—

downloads a day

What it does

Causal memory and tamper-evident audit log for AI agents: log, why, search, time travel.

Quickstart

# 1 — add it to Claude Code
claude mcp add --scope user zizkadb-mcp '--' uvx zizkadb-mcp

# 2 — ask your agent for something these tools do
#   log_event, reset_log_chain, search_memory and 10 more

ZizkaDB costs nothing on mcprush: there is no plan to choose here, no cap to set and nothing mcprush can bill you for.

Publishernot claimed

ZIZKA-AI-SL1 server

1 category

All 1 server

Collected from a public index. Nobody has claimed this account, so nothing here was written by its author — claim it if it is yours.

Installation

Where are you running it?

Every route below installs the same thing and ends at the same approval screen. This one runs on your machine: your client starts ZizkaDB as a process under your own user, with your files and your network, so the tool surface below is what it can do to your computer rather than to a server somewhere else. Its scan result is on the Security tab — read the surface before you approve it.

Claude Codedeclared by this listing

This is a public server: you run it yourself and this marketplace is not in the path. You need uv (docs.astral.sh/uv): uvx comes with it. Claude Code registers it in one command; --scope user makes it available in every project. The '--' is in quotes so that PowerShell passes it on.

claude mcp add --scope user zizkadb-mcp '--' uvx zizkadb-mcp

Reconnect, or start a new session, and the tools appear in the model’s tool list.

Free on mcprush, and it runs from its publisher’s own package: this marketplace is not in the path, so there is no key of ours to hold and no card to add.

What it asks of you

Nothing is required, but it can take ZIZKADB_HOST (ZizkaDB API URL, e.g. http://localhost:8000 for self-hosted. Default: ZizkaDB Cloud.), ZIZKADB_API_KEY (Agent API key (zizkadb_live_...). Not needed for a local development install.), ZIZKADB_TELEMETRY (Set to false to turn off the anonymous install ping.) if you want to set them.

What it writes

One config entry your client uses to start the process locally. A local server runs with your file system and your network, which is why it is priced without metering.

What you approve

13 tools, with what each one reads, writes and reaches shown before you agree — the same list on every route above. Read the tool surface.

Tool surface

What the model actually sees. Descriptions are diffed on every release — see version history.

Only tools that write or reach the network

Read from its source code. A server that builds its tool list when it starts can list more tools than this.

Log an event to ZizkaDB.

Parameter	Type	Required	Notes
agent	string	yes	Unique identifier for your agent (e.g. "support-bot"): letters, numbers, dots, dashes and underscores, starting with a letter or number
event	string	yes	What happened (e.g. "tool_call", "user_message", "decision")
data	object	yes	Any dict of relevant data (tool name, query, result, etc.)
session_id	string	no	Groups related events into a session (optional)
parent_id	string	no	event_id of the event that caused this one -- enables causal lineage (optional)

Start a fresh causal chain (new session_id, clear parent_id).

Takes no parameters.

Search over all logged agent events.

Parameter	Type	Required	Notes
query	string	yes	Natural language description of what you're looking for
agent	string	no	Filter to a specific agent (optional -- leave blank to search all)
limit	integer	no	Number of results (default 10, max 50)
mode	string	no	Leave blank for the server default, or "keyword", "semantic", "hybrid"

Prove an agent's history is untouched.

Parameter	Type	Required	Notes
agent	string	yes	The agent whose history to check

Get a formatted memory context block ready to inject into a system prompt.

Parameter	Type	Required	Notes
agent	string	yes	Your agent identifier
task	string	yes	What the agent is about to do (guides semantic retrieval)
max_tokens	integer	no	Token budget for the context block (default 2000)
session_id	string	no	Current session ID to exclude from context (optional)

Trace the causal chain that led to an event.

Parameter	Type	Required	Notes
event_id	string	yes	The event you want to explain
depth	integer	no	How many levels back to trace (default 10)

List recent events for an agent.

Parameter	Type	Required	Notes
agent	string	yes	Agent identifier to query
limit	integer	no	Number of events to return (default 20)
event_type	string	no	Filter by event type, e.g. "tool_call" (optional)

Replay the exact state of an agent at a specific point in time.

Parameter	Type	Required	Notes
agent	string	yes	Agent identifier
timestamp	string	yes	ISO 8601 datetime, e.g. "2026-05-10T15:00:00Z"

Behavioral baseline and drift score for an agent.

Parameter	Type	Required	Notes
agent	string	yes	—
recent_window	integer	no	—
window	string	no	—

Token usage and cost report for an agent over a date range.

Parameter	Type	Required	Notes
agent	string	yes	—
from_	string	yes	—
to	string	yes	—
granularity	string	no	—

Deterministic token optimization suggestions from real usage events.

Parameter	Type	Required	Notes
agent	string	yes	—
from_	string	no	—
to	string	no	—
granularity	string	no	—

Summarise what happened in a session.

Parameter	Type	Required	Notes
session_id	string	yes	The session to analyse

Delete all events matching a filter (GDPR right to erasure).

Parameter	Type	Required	Notes
filter_key	string	yes	Field in event data to match (e.g. "user_id", "email", "session_id")
filter_value	string	yes	Value to match (e.g. "user_123", "alice@example.com")

Free

$0 forever

Every tool, no call limit
No card, and mcprush charges nothing for it
Source published under MIT
Runs on your machine — nothing of it reaches our gateway
Nothing to cap, because mcprush bills nothing

Nothing here is billable: mcprush charges nothing to install ZizkaDB or to call it, at any volume.

ATrust score 96 / 100

Release 0.1.12 · scanned 1 Nov 2026

Two independent axes, because powerful and malicious are different questions. The grade is threat only. The capability level is blast radius, and it is never a penalty on the grade — it is priced as one subtract-only term in the score, where you can see it.

Threat score100

Client score96

Findings1 threat · 0 capability

Capabilitymoderate

Coveragesource

Verificationrepo

Surface digestd939bfc1163f7a19229a2ae8dd82623435d9f834cc09fbf9a14e8a5ed4e6589e

Score vectorthreat 100 → client 96

Term	Level	What it prices	Points
capability-exposure	moderate	capability blast radius (moderate) — client exposure if the model is manipulated	−3
verification-discount	repo	publisher verification (public source) — no provenance, but the source is public and inspectable	−1
coverage-honesty	source	inspection depth (source) — how much of the target the scan could see	−0

score = clamp(round(threatScore − E_cap − E_ver − E_cov), 0, 100). No model, no opinion, no network call at scoring time: the same methodology version against the same bytes always produces this number, and every line above is a field of the report rather than a summary of it.

What the scan could actually read

A grade is only as meaningful as its coverage, so the scanner publishes its own depth before it publishes its result.

Tool surfaceread

Implementation sourceread

Package metadataread

Live transportnot available

Tools were statically extracted from the published source (13 recovered), not enumerated from a running server. Tool-poisoning, Unicode-smuggling, capability and toxic-flow analysis ran on this inferred surface, but a mis-parsed registration could be missed or mis-attributed, so tool-derived findings are capped below “confirmed”. To grade the real runtime surface, scan the running server: --command "npx -y <package>".

Capability — what it could do if the model were manipulated

Tags derived from each tool’s schema and the implementation, not from what the tool calls itself. moderate is the level these add up to.

Tool	Capability tags	Why the tag was assigned
log_event	no tags
reset_log_chain	no tags
search_memory	no tags
verify_chain	no tags
get_context	no tags
why	no tags
query_events	no tags
time_travel	no tags
get_baseline	no tags
get_token_usage	sensitive-source	sensitive-source: keyword "get_token"
get_token_optimization	sensitive-source	sensitive-source: keyword "get_token"
memory_diff	no tags
forget	no tags

Toxic-flow graph

The lethal trifecta, checked as a graph rather than as a checklist: untrusted input, a sensitive source and an external sink have to meet before there is a path worth worrying about.

The public result for this release does not print the flow graph, so there is nothing to show here. That is not the same as "no paths were found": what the scan did read is above, under coverage.

Supply chain and provenance

Integrityfirst-seen

Surface digestd939bfc1163f7a19

Scanned2026-11-01

This is the first scan of this surface here, so there is nothing yet to compare it against.

Every result on this tab comes from one deterministic pass over the published package — offline, rule by rule, and auditable line by line above. Same methodology version, same bytes, same score.