ztrackBLOCK
ztrack: evidence-gated task tracking for AI agents
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
ztrack
Done is earned, not declared. A verification gate for AI coding agents — every “done” backed by a real commit and proof, or it doesn’t pass.
Setup · Usage · Guide · Support · Startup Pilot
green; fake SHA -> exit 1" width="680">
AI coding agents close tickets on prose. "All tests pass, feature complete" — and the commit it cited never existed. Your tracker stored the claim with perfect fidelity and verified nothing.
ztrack is a typechecker for your issue tracker. A checked acceptance criterion must cite a commit SHA that exists in git, plus the eviden
16fe03c74cdeOBSERVED · 2026-10-12Connect
Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control. Replace the environment placeholders with a token scoped to the least it needs.
claude mcp add ztrack-visualizer --env GH_TOKEN=${GH_TOKEN} --env GITHUB_TOKEN=${GITHUB_TOKEN} -- npx -y [email protected]{
"mcpServers": {
"ztrack-visualizer": {
"command": "npx",
"args": [
"-y",
"[email protected]"
],
"env": {
"GH_TOKEN": "${GH_TOKEN}",
"GITHUB_TOKEN": "${GITHUB_TOKEN}"
}
}
}
}Exposed tools (7)
4 read · 3 write · 0 destructive.
| Tool | Risk | Description |
|---|---|---|
tracker_check | read | Validate the tracker through the single pipeline (parse → strict schema → pure rules: state gates, evidence/SHA anchoring). Returns { ok, summary, findings }; ok=false means findings must be resolved with evidence. |
tracker_fmt | write | Canonicalize an issue body (whitespace, checkbox markers, section order). write=false previews. |
tracker_init | read | Initialize ztrack in the current project (writes .volter/tracker-config.json with the selected validation preset, day-one check defaults, and a managed .gitignore). Call this first in a fresh repo — the server starts without a config so an MCP-only agent can bootstrap. Idempotent. |
tracker_issue_create | write | Create an issue. |
tracker_issue_list | read | List tracker issues (state filter optional). |
tracker_issue_view | read | View one issue including its body. |
tracker_patch | write | Edit an issue by overlaying a typed fragment onto the issue (or one acceptance criterion, by acId) and re-serializing through the active preset. |
Trust audit
BLOCKgrade D · trust 61/100 Do not install this without reading the findings. The audit found something that could harm you or your machine.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | FAIL |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | PASS |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- none-observed
- Network
- declared (4 observation(s))
- Shell
- declared (5 observation(s))
- Dependencies
- not all pinned
- Secrets in source
- none-found
Findings (18)
const list = JSON.parse(await exec(['list', '--json'])) as Json[];
const run = (args: string[]) => exec(args).then(() => undefined);
const out = JSON.parse(await exec(args)) as Json;
if bun -e "try { process.exit((await fetch('http://127.0.0.1:$port/')).status === 200 ? 0 : 1); } catch { process.exit(1); }" >/dev/null 2>&1; then up=1; break; ficonst base = `http://127.0.0.1:${port}`;.termfleet-virtual-tmux-oasd-layout.json
.termfleet-virtual-tmux-p10-layout.json
.termfleet-virtual-tmux-proof1-layout.json
.termfleet-virtual-tmux-termfleet-virtual-layout.json
import { assertAcSubLineFidelity, assertNotePositionFidelity, assertRoundTripFidelity, assertSdlcGrammarConformance, assertVisualizerSpecConformance } from '../../src/testkit/presetConformance.ts';import { assertAcSubLineFidelity, assertNotePositionFidelity, assertRoundTripFidelity, assertSdlcGrammarConformance, assertVisualizerSpecConformance } from '../../src/testkit/presetConformance.ts';import { assertRoundTripFidelity, assertVisualizerSpecConformance } from '../../src/testkit/presetConformance.ts';import { assertReadOnlyRoundTripExemption, assertVisualizerSpecConformance } from '../../src/testkit/presetConformance.ts';import type { CoreRoot, Preset } from '../../src/core/engine.ts';@types/bun, @types/node, @types/react, @types/react-dom, @volter/twin-github, @volter/twin-jira, @volter/world-core, happy-dom
react, react-dom, @types/react, @types/react-dom
claude -p 'Adopt ztrack in this repository. Start by reading https://github.com/volter-ai/ztrack/blob/main/README.md, https://github.com/volter-ai/ztrack/blob/main/docs/GUIDE.md, https://github.com/vo
A read-only web view (requires [Bun](https://bun.sh); first run installs its client deps once). It
Gates applied: no_behavioural_pass.
16fe03c74cdefull audit observations/trust-audit/mcp-server/volter-ai__ztrack.json · Report an issue / request a re-scanAudit history
Every audit this server has had. A grade with a past is a grade somebody is still checking.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-12 | 16fe03c74cde | BLOCK | D | 61 | first audit |
Questions
What is the ztrack MCP server?
ztrack: evidence-gated task tracking for AI agents
What tools does ztrack expose?
7 in total: 4 read-only, 3 that write, and 0 that can delete or overwrite. Every one is listed on this page with its risk.
Is ztrack safe to connect to an agent?
No — not without reading the findings first. The audit graded it D (61/100) and found 3 critical or high issues in the source. Each one is listed on this page with the file and line it is on.
What credentials does ztrack need?
It reads GH_TOKEN and GITHUB_TOKEN from the environment. Give it a token scoped to the least it needs — an agent that can be talked into calling a tool can be talked into calling it with your credentials.
How current is this page?
The grade is for one exact copy of the source (16fe03c74cde), read on 2026-10-12. The repository is watched and re-audited when it changes.