Pack local or remote codebases into AI-friendly files that LLMs and coding agents can read or search
Nothing to run from here
knock 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
7
downloads a day
local
runs on your machine
89
downloads in 30d
What it does
Code map and memory for AI coding agents: key functions, line ranges, callers, why code changed.
Quickstart
# 1 — add it to Claude Code
claude mcp add --scope user knock-mcp '--' uvx knock-mcp serve --repo . --toolset map
# 2 — ask your agent for something these tools do
# get_context, search_knowledge, get_knowledge and 8 more
knock costs nothing on mcprush: there is no plan to choose here, no cap to set and nothing mcprush can bill you for.
Publishernot claimed
HAAll 1 server
hasangilani1 server
1 category
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 knock 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 knock-mcp '--' uvx knock-mcp serve --repo . --toolset mapReconnect, 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 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
11 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.
Read from its source code. A server that builds its tool list when it starts can list more tools than this.
Use before exploring (how/why questions, unclear bugs, unfamiliar code); skip for small edits one grep can locate. Returns a compact, ranked context package for a natural-language task: relevant knowledge items (with source references, provenance and freshness), matching symbols with file:line,…
| Parameter | Type | Required | Notes |
|---|---|---|---|
| task | string | yes | What you are trying to do or understand, e.g. 'add rate limiting to the login endpoint'. |
| paths | array | no | Optional repository-relative files you already know are relevant; their knowledge is included first. |
| max_chars | integer | no | Response budget in characters (default from config, ~8000). |
| repository | no | — |
Full-text search over stored knowledge items, ranked by relevance, source authority, verification and freshness. Filter by category and/or path prefix. Returns concise facts with source references; use get_knowledge for full provenance. Categories: ….
| Parameter | Type | Required | Notes |
|---|---|---|---|
| query | string | yes | Keywords or a question, e.g. 'how are database migrations run'. |
| categories | array | no | Only return these categories. |
| path_prefix | string | no | Only items citing files under this repository-relative path, e.g. 'src/auth'. |
| include_inactive | boolean | no | Also return superseded/invalidated items (for history). |
| limit | integer | no | Maximum results (default 10). |
| max_chars | integer | no | Response budget in characters (default ~4000). |
| repository | no | — |
Fetch one knowledge item by id with full provenance: every source reference and whether it still matches the indexed file, lifecycle status and reason, verification, timestamps, revision, and links (supersedes/conflicts). Set include_evidence=true to also get the cited source excerpts.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| knowledge_id | string | yes | Item id, e.g. 'k_1a2b3c4d5e6f'. |
| include_evidence | boolean | no | Include current excerpts of the cited source lines. |
| repository | no | — |
Read the current, authoritative contents of an indexed repository file (line-numbered, secrets redacted), by line range or by symbol name. Use it to verify knowledge before relying on it. Only files in the index are served: excluded, ignored, sensitive or out-of-repository paths are refused. Max…
| Parameter | Type | Required | Notes |
|---|---|---|---|
| path | string | yes | Repository-relative file path, e.g. 'src/app/server.py'. |
| start_line | integer | no | First line (1-based). |
| end_line | integer | no | Last line (inclusive). |
| symbol | string | no | Instead of lines: a function/class/method name defined in this file. |
| repository | no | — |
Persist one reusable, verified-by-you fact for future sessions: a debugging root cause and fix, a non-obvious workflow step, a decision and its rationale, a convention, or where something is implemented. One concise fact per call; cite the files (and lines) that support it so it is automatically…
| Parameter | Type | Required | Notes |
|---|---|---|---|
| category | string | yes | Knowledge category. |
| content | string | yes | The fact, stated concisely and specifically (max ~2000 chars). |
| title | string | no | Short title (optional). |
| rationale | string | no | Why it is so / why it was decided (optional). |
| sources | array | no | Supporting files: [{"path": "src/x.py", "start_line": 10, "end_line": 40}]. Paths must be indexed files. |
| entities | array | no | Key names this concerns (functions, services, config keys). |
| confidence | number | no | Your confidence 0..1. |
| supersedes | array | no | Ids of existing items this replaces (they become 'superseded'). |
| repository | no | — |
Change an item's lifecycle after checking it against the source: 'verify' (agent-recorded items only; marks it verified at the current indexed revision and clears staleness) or 'invalidate' (it is wrong or no longer applies; kept for history but excluded from results). Give a short reason.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| knowledge_id | string | yes | Item id. |
| action | string | yes | 'verify' or 'invalidate'. |
| reason | string | no | Why (recorded with the item). |
| repository | no | — |
Report index freshness (indexed vs current Git revision, files modified since indexing, stale items, last run), knowledge counts by category/status/provenance, cache statistics and local usage counters.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| repository | no | — |
Incrementally re-index the repository: only added/modified/deleted files are reprocessed (deterministic, local, no model calls). Knowledge from changed sources is updated or invalidated, and agent-recorded items citing changed files are marked stale. Use full=true only if the index seems…
| Parameter | Type | Required | Notes |
|---|---|---|---|
| full | boolean | no | Rebuild every file instead of only changed ones. |
| repository | no | — |
Use instead of exploring: how/why questions, unclear bugs, multi-module changes, code you can't find with one search. Skip it for small edits one grep can locate. Returns the code relevant to a task as a compact structural map: functions and classes with signatures, exact line ranges, what they…
| Parameter | Type | Required | Notes |
|---|---|---|---|
| task | string | yes | The question or change, in your own words, e.g. 'add a debug log when session cookie verification fails'. |
| repository | no | — |
Exact, line-numbered code for a symbol ('open_session', 'SecureCookieSessionInterface.open_session') plus what it calls and who calls it; or a line range ('src/app.py:120-160'); or a file's skeleton ('src/app.py'). Only indexed files; secrets redacted.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| target | string | yes | Symbol name, 'path:start-end', or a file path. |
| repository | no | — |
Save one non-obvious fact a future session would need (root cause, hidden constraint, why something is done a certain way). Cite files so it is flagged when they change. Skip anything the code or docs already say.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| note | string | yes | The finding, one or two sentences. |
| files | array/string | no | Supporting files, e.g. ['src/app.py:120-160'] or 'src/app.py'. |
| repository | no | — |
Free
$0 forever
- Every tool, no call limit
- No card, and mcprush charges nothing for it
- Source published under Apache-2.0
- 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 knock 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 0.2.0 · scanned 3 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
Findings3 threat · 0 capability
Capabilityhigh
Coveragesource
Verificationsource
Surface digestc8d4c229e9f7ebedc77e2fb1411e9e1dcabc9d90efa0b68ae14962e090ef5135
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 surfaceread
Implementation sourceread
Package metadataread
Live transportnot available
Tools were statically extracted from the published source (11 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>".
8 of the 11 recovered tools are registered only when toolset == "map" is false (src/knock/mcp_server.py:62) — typically when a key or option is configured. They are part of the server's surface and were analysed; a server started without that configuration exposes only the other 3.
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.
3 tools are registered unconditionally · 8 more only when toolset == "map" is false (src/knock/mcp_server.py:62) — typically when a key or option is configured. All 11 are in the table below and were analysed; the conditional ones are marked.
| Tool | Capability tags | Why the tag was assigned |
|---|---|---|
| get_context conditional | no tags | |
| search_knowledge conditional | no tags | |
| get_knowledge conditional | untrusted-input | untrusted-input: keyword "fetch" |
| get_source conditional | no tags | |
| record_knowledge conditional | no tags | |
| update_knowledge_status conditional | no tags | |
| get_repository_status conditional | no tags | |
| refresh_repository conditional | no tags | |
| map | no tags | |
| open | no tags | |
| remember | 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 digestc8d4c229e9f7ebed
Scanned2026-11-03
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 0.2.0. Pin a release below and the command asks for exactly that one.
0.2.0
3 Nov 2026
latestNo release note was published with this version.
0.1.1
undated
- The version this listing was on when mcprush began following its releases. The registry was not asked when it shipped, so this row carries no date.
—
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 knock 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.