Query Markup Documents - On-device hybrid search for markdown files with BM25, vector search, and LLM reranking
Nothing to run from here
witness is installed from its publisher's own source and answers where it runs, so this marketplace is not in the path of a single call. There is no address here to send one to, and a panel that pretended otherwise would be showing you an answer we made up. Install it and call it from your own client — the Installation tab has the entry for each one.
94/100
trust score
70
downloads a day
local
runs on your machine
2k
downloads in 30d
What it does
Distills your sessions or any text corpus into a dated history of how its subject changed.
Quickstart
Install it the way its author documents:
1. Claude Code: git clone https://github.com/IngTian/witness && cd witness && ./install.sh claude
2. OpenCode: npm i -g @witness-ai/opencode — then add the plugin to opencode.json
3. Codex: ./install.sh codex from the same checkout, then approve the hooks
4. with /hooks inside Codex (they stay inert until you do)
5. Windows: unpack witness-v0.10.0-windows-<arch>.zip from Releases,
6. then .\witness.exe wire claude
The full guide: https://github.com/IngTian/witness#install
Publisher
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 witness 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.
Install it the way its author documents:
1. Claude Code: git clone https://github.com/IngTian/witness && cd witness && ./install.sh claude
2. OpenCode: npm i -g @witness-ai/opencode — then add the plugin to opencode.json
3. Codex: ./install.sh codex from the same checkout, then approve the hooks
4. with /hooks inside Codex (they stay inert until you do)
5. Windows: unpack witness-v0.10.0-windows-<arch>.zip from Releases,
6. then .\witness.exe wire claude
The full guide: https://github.com/IngTian/witness#install
Tool surface
What the model actually sees. Descriptions are diffed on every release — see version history.
Delete one observation from the archive by obs_id (get ids from search_observations). Use to prune a wrong or stale observation; the profile re-derives from the remaining ones on the next review. The profile (facets) itself is not directly editable — observations are the input layer.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| obs_id | string | yes | the obs_id to delete (get ids from search_observations) |
Read the synthesized growth facets (structured attributes) for a lens. Returns current facet values — trajectory arcs, mastery markers, resilience patterns. Defaults to the always-on 'default' lens (cross-domain); pass lens='math' etc. for a repo lens. (Facets are per-lens, so there is no cross-lens 'unified' facet set — that aggregate exists only for the narrative; use get_profile for it.) For a
| Parameter | Type | Required | Notes |
|---|---|---|---|
| lens | string | no | which lens to read; omit for the always-on 'default' lens (cross-domain), or name a repo lens like 'math' |
Read the narrative growth profile (readable prose) for a lens. Omit lens (or pass 'unified') for the cross-lens portrait — an aggregate blending ALL active lenses, not a lens itself. Pass lens='default' for the always-on lens's OWN narrative (distinct from the unified blend), or lens='math' etc. for another lens. (This defaults to the unified aggregate, whereas search_observations/get_facets defau
| Parameter | Type | Required | Notes |
|---|---|---|---|
| lens | string | no | omit for 'unified' (the cross-lens portrait blending ALL lenses — an aggregate view, not a lens); or name one lens: 'default' (the always-on lens's own narrative), 'math', etc. |
Record a decision-aware observation about the user's growth/change in THIS session. Use when you (the assistant) notice something worth tracking that you have context for now but a later reviewer would miss. Stored verbatim. Tag with the relevant lens.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| session | string | yes | the current session_id |
| lens | string | no | lens tag (default 'default') |
| dimension | string | yes | which axis this observation belongs to |
| observation | string | yes | the noticed growth/change, one sentence |
| evidence | string | no | short anchor to what happened |
| poignancy | integer | no | 1-10 salience (default 5) |
Search the user's growth archive for observations related to a query. Returns past observations about how the user thinks/works/changes. Defaults to the 'default' lens (cross-domain); pass lens='math' etc. for a repo-scoped view.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| query | string | yes | the growth/pattern question to search for |
| lens | string | no | which lens to search; omit for the always-on 'default' lens (cross-domain), or name a repo lens like 'math' |
| k | integer | no | max results (default 8) |
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
What counts against your monthly calls
| Tool | Unit | Calls used | Out of the allowance |
|---|
Nothing here is billable: mcprush charges nothing to install witness or to call it, at any volume.
Call allowance · this installenforced at the gateway
Monthly allowance, in calls
—
—
Alert at 80% · then the tool returns a normal MCP error
Used this month: —
ATrust score 94 / 100
Release 0.12.0 · scanned 6 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 score94
Findings2 threat · 0 capability
Capabilityhigh
Coveragesource
Verificationsource
Surface digest9b81f6cec6154f0e3dcea4ad91293314f170ea0fd95a888ce905dff1740b829b
Score vectorthreat 100 → client 94
| Term | Level | What it prices | Points |
|---|---|---|---|
| capability-exposure | high | capability blast radius (high) — client exposure if the model is manipulated | −6 |
| verification-discount | source | publisher verification (provenance) — cryptographic build provenance ties the artifact to its source | −0 |
| 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 surfacenot available
Implementation sourceread
Package metadataread
Live transportnot available
No tools were enumerated, so prompt-injection, capability and toxic-flow analysis had no tool surface to inspect. To grade a package’s real runtime tools, 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. high is the level these add up to.
| Tool | Capability tags | Why the tag was assigned |
|---|---|---|
| No per-tool rows for this release — see the note below. | ||
can execute shell commands or code
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 digest9b81f6cec6154f0e
Scanned2026-11-06
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.
Release history
Always latest — the install command names no version, so it installs the newest release when it runs; the current release on file is 0.12.0. Pin a release below and the command asks for exactly that one.
0.12.0
6 Nov 2026
latestNo release note was published with this version.
0.11.0
2 Nov 2026
No release note was published with this version.
0.10.0
1 Nov 2026
- Codex sessions now feed your archive with one command instead of hand-edited JSON.
—
no reviews yet
Who can review
One review per account, from a signed-in account that does not publish this listing. Nothing else is asked — you can say what you think before you install it. Publishers can reply once per review, and a review can be edited or deleted by whoever wrote it.
Sign in to write one
Any signed-in account that does not publish witness can review it, once — before installing it or after. A review from an account that has installed it is marked as one.
No reviews yet
Nobody has reviewed this listing. The rating on the card is the mean of the reviews written here and nothing else, so there is no rating until somebody writes the first — which takes a signed-in account that does not publish it, and nothing else.