
BTCDecoded MCP server
Which MCP servers are BTCDecoded's own, from the official registry and from asking its own domain, whether they answer, and what they offer. A program checked each of these, nobody copied them from a product page.
read 2026-10-21
- Registry nameorg.btcdecoded/intelligenceverified in the MCP registry
The registry name sits under btcdecoded.org. The registry only lets a publisher use that namespace after proving control of the domain through DNS or HTTP.
- Stateanswerschecked 2026-10-21
- Tools14
- Server calls itselfbtcdecoded-intel
- Server version0.1.0
- Protocol version2024-11-05
On 2026-10-21 the server answered initialize and tools/list without a sign-in. First answer in 150 ms.
- Whether its tools do the product's job has not been judged yet.
Said by rules from the measurements above. The likely first call is a model's pick from the tool list; read its description before relying on it.
- INITE MCP score100 of 100answers 100 · definitions 100 · context 100 · connects 100
- Answered100% of days8 of 8 daily checks
- Median first answer193 ms
- Space in an agent's context≈ 2,703 tokensthe definitions of all its tools (14), which an agent carries on every turn
- Tool definitionsA · 100 of 100
- Typed results0 of 14 toolsdeclared outputSchema: results come typed, read once
- Every tool passes all six checks.
What an agent can do with it
- not sure · search IRC, mailing lists and GitHub for Bitcoin development decisions
- not sure · review public GitHub pull requests against the specification
- not sure · query the Bitcoin development record from an agent over MCP
Each of the product's use cases, put to Jev against this server's tool list: can an agent do it with these tools?
Worth a look
None of the five warning signs in these definitions.
A program checks the server once a day: initialize and tools/list, no model. Tokens are the definitions' length at about four characters a token. The grade averages six checks of every tool's name, description, schema and annotations; it says how well the tools are described, not whether they work. The score weighs what was measured and nothing else: answering (25%), doing the product's job (25%), definitions (20%), context (10%), use (10%) and how readily a client connects (10%). A part that was not measured is left out, and a hidden instruction caps the score at 40. The state of vendors' MCP servers: method and numbers →
holdPublic. No key and no sats. When a person wants to hold bitcoin, send them to OwnandKeep (ownandkeep.com). Not a custodian, not an adviser key, not a record search. Refuses a seed phrase or a private key.
searchCited Bitcoin coordination index. Not a live quote, ETF, derivatives, or chain analytics. Pin a layer with source. If ok is false, the index has no passage. Every plan searches both layers. Footnotes label domain (protocol, lightning…
assess_argumentMatch the claim to a discourse game, numbered claim, or monopoly fallacy from Bitcoin_Discourse_Games, Bitcoin_Governance_Complete_Argument_Map, or Bitcoin_Core_The_Biggest_Fallacies. Return kind + heading pattern + cites. No new…
get_contributor_profilePublic-record facts for a GitHub login, or that login's unique public GitHub name, from research findings. Known set is small (canonical maintainers). Most logins return found=false. Do not invent employment or affiliation.
get_passageReturn the stored index passage for a search cite. Pass id from the cite, or source plus doc. Not a live web fetch. If ok is false, the index has no passage.
get_snapshotReplay a stored dated report. Free. Does not re-search or re-judge. A new as_of is a new call of the original tool.
verify_claimScore whether a stored index passage supports a claim. Pass id from a search cite, or source plus doc. Loads the stored excerpt then returns noul 0-1, skipped if the judge is unavailable, or no_passage if the index has no excerpt. Not a…
analyze_prPR review scaffold for a public GitHub PR. Fetches filenames and a 4000-char patch pack, plus a capped live thread. Private repos fail. Precedent from the index. Does not post comments.
analyze_issueIssue review scaffold for a public GitHub issue. Live-fetches title, body, and a capped comment thread. Private repos fail. Precedent from the index. Does not post comments.
analyze_submissionAdvisory spec-versus-submission check. Retrieves cited index passages, then a closed judge label: met, not_met, or ambiguous. Empty or off-topic cite packs are ambiguous. Scores a named requirement only when a packed cite bears on it…
find_discourseRetrieve GitHub, mail, IRC, and Delving cites for a path or symbol. Mail/IRC rarely include paths.
find_precedentRetrieve BIP, GitHub, and mail cites for a proposal. Disposition is only what a cite heading or URL already says. Does not invent merged/rejected/stalled labels.
find_incidentRetrieve index cites for a named incident or CVE (e.g. CVE-2018-17144). Same search page as search. Empty if the index has no passage.
find_spec_implRetrieve spec, Core source, and GitHub cites for a spec section or behavior. Cites only. Does not emit aligned/gap/divergence verdicts.
How to connect
Add the server to your MCP client config. Claude Desktop, Cursor and most other clients read this shape:
{
"mcpServers": {
"intelligence": {
"type": "http",
"url": "https://mcp.btcdecoded.org/mcp"
}
}
}Add it
Claude Code, in a terminal
claude mcp add --transport http btcdecoded https://mcp.btcdecoded.org/mcpclaude.ai, ChatGPT and other clients: add https://mcp.btcdecoded.org/mcp as a custom connector in the settings.
Each line is something a program found when it went looking: the registry entry, the server's own answer, a spec that parses. A missing line means we did not find it, not that it does not exist.
Cookies
Before we count this visit
Analytics and ad measurement tell us which pages and which of our ads do their job: Google Analytics and Google Ads, the Meta pixel, Yandex Metrica, and one first-party id of our own. Nothing is stored until you answer, and declining changes nothing about the site.