Multiplayer mode for AI coding agents: team huddles, master prompt, task board, file claims.
Nothing to run from here
Tokesave 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.
93/100
trust score
1
downloads a day
local
runs on your machine
71
downloads in 30d
What it does
The ultimate AI token-saving MCP middleware — compression, caching, injection shield, multi-server routing, and more.
Quickstart
# 1 — add it to Claude Code
claude mcp add --scope user tokesave-mcp '--' npx -y tokesave-mcp
# 2 — ask your agent for something these tools do
# auto_setup, compress_text, compress_batch and 24 more
Tokesave costs nothing on mcprush: there is no plan to choose here, no cap to set and nothing mcprush can bill you for.
Publishernot claimed
IRAll 1 server
irsyadmr3541 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 Tokesave 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 Node.js — the current LTS (24) is the safe choice; npx 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 tokesave-mcp '--' npx -y tokesave-mcpReconnect, 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
27 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.
CALL THIS FIRST on every new session. Prints optimal tool routing rules for this session + current config status. Ensures TokeSave is active and correctly configured.
No parameter schema on file.
USE THIS before sending any long text (>200 words) to save tokens. Strips filler words, abbreviates common terms, removes articles. Pass filename for code files to enable AST skeleton mode.
No parameter schema on file.
USE THIS when you need to read multiple files at once. Single tool call instead of N calls. Compresses all files concurrently and returns results together.
No parameter schema on file.
USE THIS when you only need to understand a file, not quote it verbatim. Returns file content within a token budget — auto-escalates compression mode to fit. Ideal for large files where full content is not needed.
No parameter schema on file.
USE THIS instead of reading an entire file when you only need one function or class. Extracts just the target function — reads only what
No parameter schema on file.
USE THIS when comparing file versions. Returns only changed lines instead of both full files — saves tokens proportional to file size.
No parameter schema on file.
USE THIS instead of fetching URLs directly. Fetches + strips HTML boilerplate (scripts, styles, nav, ads) + compresses. Returns clean readable text at fraction of original token cost.
No parameter schema on file.
USE THIS for very large files (>50KB) that need to be processed in parts. Splits compressed content into chunks of specified token size.
No parameter schema on file.
Search previously read/compressed content by semantic similarity. USE THIS before re-reading files — if content was compressed before, this returns it instantly without disk I/O or re-compression.
No parameter schema on file.
Scan external content (web pages, tool outputs, user input) for prompt injection attempts before processing. Returns sanitized text.
No parameter schema on file.
Change compression intensity. lite=minimal, standard=no articles, aggressive=abbreviations (default), brutal=symbols, oblivion=vowel removal. Higher = more savings but less readable.
No parameter schema on file.
Get token savings statistics for current session and lifetime.
No parameter schema on file.
Get current TokeSave state: active compression mode, which features are enabled, session stats.
No parameter schema on file.
Create a named virtual agent with a role for task delegation tracking. NOTE: multi-agent delegation increases total token usage — do not use this for token-saving purposes.
No parameter schema on file.
Assign a task to an existing swarm agent. NOTE: this triggers additional model calls per agent — not a token-saving tool.
No parameter schema on file.
Get status of all swarm agents and their current tasks.
No parameter schema on file.
Test all 5 compression modes on text/file and compare savings, readability, info loss. Returns recommended mode.
No parameter schema on file.
Scan git repo for token waste: large files, binaries, generated code. Estimates potential token savings.
No parameter schema on file.
Analyze repo + generate .gitignore + list files to remove from tracking. Detects project type automatically.
No parameter schema on file.
Apply the cleanup: generate .gitignore + git rm --cached for binaries/generated files. Must commit separately.
No parameter schema on file.
Search files for pattern, return only matching lines with context. Avoids reading full files. Scans up to 500 files, caps at 200 matches to prevent hang on large repos.
No parameter schema on file.
Read specific line range from a file. Avoids reading huge files entirely just for a few lines.
No parameter schema on file.
Lightweight directory tree with file sizes and token estimates. No file content read.
No parameter schema on file.
Detect existing MCP servers from Cursor/Claude/Kiro configs and generate tokesave.config.json with servers:{} auto-filled. Prints instructions for updating mcp.json.
No parameter schema on file.
Check session context window pressure based on total tokens processed. Returns recommended compression mode and auto-escalates if pressure is high.
No parameter schema on file.
Compress a long AI response for storage/reference in multi-turn conversations. Optionally stores compressed summary in vector memory for later recall.
No parameter schema on file.
No parameter schema on file.
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 Tokesave 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 93 / 100
Release 1.0.5 · scanned 15 Oct 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 score93
Findings3 threat · 1 capability
Capabilityhigh
Coveragesource
Verificationrepo
Surface digest51c042cc4a8f9f626ea3483fbeb57c2bee6801d3f6ceb7d2c51182c5f8a4776b
Score vectorthreat 100 → client 93
| Term | Level | What it prices | Points |
|---|---|---|---|
| capability-exposure | high | capability blast radius (high) — client exposure if the model is manipulated | −6 |
| 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 (27 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>".
27 of the 27 recovered tools are registered only when args[0] === 'marketplace' is false (src/index.js:142) — 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 none of the recovered tools.
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.
No tool is registered unconditionally · 27 only when args[0] === 'marketplace' is false (src/index.js:142) — typically when a key or option is configured. All 27 are in the table below and were analysed; the conditional ones are marked.
| Tool | Capability tags | Why the tag was assigned |
|---|---|---|
| auto_setup conditional | no tags | |
| compress_text conditional | no tags | |
| compress_file conditional | no tags | |
| compress_batch conditional | no tags | |
| summarize_file conditional | no tags | |
| read_function_body conditional | no tags | |
| diff_files conditional | no tags | |
| compress_url conditional | no tags | |
| compress_file_chunked conditional | no tags | |
| recall_memory conditional | no tags | |
| scan_injection conditional | no tags | |
| set_mode conditional | no tags | |
| get_stats conditional | sensitive-source | sensitive-source: keyword "get_token" |
| get_status conditional | no tags | |
| swarm_create_agent conditional | no tags | |
| swarm_assign_task conditional | no tags | |
| swarm_get_status conditional | no tags | |
| benchmark conditional | no tags | |
| repo_audit conditional | no tags | |
| repo_cleanup conditional | sensitive-source | sensitive-source: keyword "list_files" |
| repo_cleanup_apply conditional | no tags | |
| grep_files conditional | no tags | |
| read_file_range conditional | sensitive-source | sensitive-source: keyword "read_file" |
| file_mtree conditional | no tags | |
| generate_proxy_config conditional | no tags | |
| check_context_pressure conditional | no tags | |
| compress_output conditional | 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 digest51c042cc4a8f9f62
Scanned2026-10-15
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 1.0.5. Pin a release below and the command asks for exactly that one.
1.0.5
undated
latestNo release note was published with this version.
—
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 Tokesave 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.