Privacy-first local MCP server: redact secrets, decode JWTs, estimate LLM costs, split long documents and check AI output for hallucinations — all on your own machine, with no network calls.
Inferred from the transports this listing declares (stdio). A client not listed here hasn’t been ruled out — it just isn’t something Forge can confirm.
Verification confirms publisher identity (repo ownership), not code safety. The security scan covers known CVEs and suspicious install scripts.
Read out of the source npm actually ships, at scan time. The package was never executed. Tools registered dynamically at runtime, or hidden inside bundled or minified code, can be missed — so this is a floor on the tool surface, not a complete census of it.
redact_sensitive_dataMask emails, API keys, credit cards, IBANs, IP addresses and phone numbers in a block of text. Use this BEFORE sending user-supplied text, logs, config files or pasted documents to any external API, or before quoting them back in a shared transcript. Runs entirely locally — nothing is transmitted.Mask emails, API keys, credit cards, IBANs, IP addresses and phone numbers in a block of text. Use this BEFORE sending user-supplied text, logs, config files or pasted documents to any external API, or before quoting them back in a shared transcript. Runs entirely locally — nothing is transmitted.
No input schema was published for this tool.
decode_jwtDecode the header and payload of a JSON Web Token and report its expiry status. Does NOT verify the signature (no secret is involved, and none should be pasted anywhere). Use this instead of an online JWT decoder — a token is a live credential and should never be pasted into a website.Decode the header and payload of a JSON Web Token and report its expiry status. Does NOT verify the signature (no secret is involved, and none should be pasted anywhere). Use this instead of an online JWT decoder — a token is a live credential and should never be pasted into a website.
No input schema was published for this tool.
estimate_llm_costCompare what one request would cost across GPT, Claude and Gemini models, sorted cheapest first. Pass the prompt text, or pass input_tokens directly if you already know the count. Useful for picking a model before running a large batch job.Compare what one request would cost across GPT, Claude and Gemini models, sorted cheapest first. Pass the prompt text, or pass input_tokens directly if you already know the count. Useful for picking a model before running a large batch job.
No input schema was published for this tool.
split_for_contextBreak a long document into chunks that fit a target context window, cutting on paragraph boundaries (falling back to sentence boundaries) so no chunk ends mid-thought. Use when a file is too large to process in one pass.Break a long document into chunks that fit a target context window, cutting on paragraph boundaries (falling back to sentence boundaries) so no chunk ends mid-thought. Use when a file is too large to process in one pass.
No input schema was published for this tool.
check_output_fidelityFlag names, numbers and dates that appear in an AI-generated summary but NOT in the source text it was based on — the usual shape of a hallucination. This is a lexical heuristic, not a fact-checker: it catches invented specifics, not wrong reasoning. Use it to decide what to verify by hand.Flag names, numbers and dates that appear in an AI-generated summary but NOT in the source text it was based on — the usual shape of a hallucination. This is a lexical heuristic, not a fact-checker: it catches invented specifics, not wrong reasoning. Use it to decide what to verify by hand.
No input schema was published for this tool.
hash_textCompute an MD5, SHA-1, SHA-256 or SHA-512 digest of a string, locally. Use instead of an online hash generator when the input is a password, secret or any private value.Compute an MD5, SHA-1, SHA-256 or SHA-512 digest of a string, locally. Use instead of an online hash generator when the input is a password, secret or any private value.
No input schema was published for this tool.
6 of 6 tools published a description.
Tool names and descriptions are written by the publisher and shown verbatim as inert text. They are the strings an MCP client passes to a model, so Forge scans them for prompt-injection patterns — any finding appears with the security scan above. “Privileged” is a keyword match on the tool name, not an audit of what the tool does: a benign-sounding name can still do anything.
Privacy-first local MCP server: redact secrets, decode JWTs, estimate LLM costs, split long documents and check AI output for hallucinations — all on your own machine, with no network calls.
Linked names open Forge’s index of every entry observed exposing that tool. Browse all indexed tools.
The crawl stopped at the depth-4 limit. Anything below that level was never resolved.
The crawl stopped at the 60-package limit. The rest of the tree was never resolved.
36 more resolved packages are not drawn here (display cap: 24). Every dependency carrying an advisory is drawn regardless of the cap. Full inventory (CycloneDX SBOM)
55 declared dependencies never landed in the tree. They are missing from Forge's resolution, not from the package.
+43 more not listed. The counts by reason above cover all of them.
Not followed: peerDependencies. This tree covers runtime dependencies only, so anything those pull in was never resolved.