probbit MCPサーバー
コミット 2026年11月13日p-bit決定エンジンを使用して、AIエージェントに一貫したスタンスとルール遵守の計画を保証する、パーソナリティ、ムード、およびメモリレイヤーを提供します。
別名 BitmapAsset/probbit
- スター
- 1
- ツール
- —
- ダウンロード数(30日間)
- —
- 最新バージョン
- —
- 最終コミット
- 2026年11月13日
- ライセンス
- Apache-2.0
probbit MCPとは
p-bit決定エンジンを使用して、AIエージェントに一貫したスタンスとルール遵守の計画を保証する、パーソナリティ、ムード、およびメモリレイヤーを提供します。
プロジェクトのREADMEから生成された概要
probbit MCP server について
AIエージェントを構築する開発者は、probbitを使用して、基盤となる大規模言語モデルから独立した、一貫性のあるパーソナリティ、ムード、およびメモリレイヤーを実装できます。CPU上でシミュレートされた正確なp-bit決定エンジンを使用することで、エージェントが安定したスタンスを維持し、厳格なルールを遵守することを保証します。このレイヤーは、決定に対して正確な確率を提供するか、結果を保証できない場合には判定を拒否するため、制約が厳しい場合にモデルが推測してしまうのを防ぎます。
このMCP serverは、Claude Code、Cursor、OpenClawを含むさまざまな環境に統合できます。APIキーを必要とせず、外部サーバーにデータを送信することなく、完全にオフラインで動作します。ユーザーは、CLI、MCP server、またはPythonやRustの統合を介して操作できます。パーソナリティはペルソナファイルとシードの純粋関数であるため、開発者が基盤となるLLMを入れ替えても、エージェントのキャラクターは同一のまま維持されます。
できること
- 1エージェントのパーソナの一貫性維持使用する言語モデルに関わらず、AIエージェントが同じキャラクターと気質を維持できるようにします。
- 2厳格な運用ルールの強制推測を拒否する制約ソルバーを使用することで、AIエージェントがプライバシーや安全性の制約に違反するのを防ぎます。
- 3エージェントの決定履歴の検証エージェントのターンのシーケンスを再生および検証し、決定ログが改ざんされていないことを確認します。
プロジェクトの README から作成されています。すべてのユースケースは、その引用と照らし合わせて確認されています。
probbit MCPのインストール方法
フォーマット検証日 2026年11月10日Cursor に追加Gooseに追加Claude Codeにprobbit MCPをインストール
$ claude mcp add probbit -- npx -y probbitCursorにprobbit MCPをインストール
~/.cursor/mcp.json
{
"mcpServers": {
"probbit": {
"command": "npx",
"args": [
"-y",
"probbit"
]
}
}
}VS Codeにprobbit MCPをインストール
.vscode/mcp.json
{
"servers": {
"probbit": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"probbit"
]
}
}
}VS Code CLIにprobbit MCPをインストール
$ code --add-mcp '{"name":"probbit","type":"stdio","command":"npx","args":["-y","probbit"]}'Claude Desktopにprobbit MCPをインストール
claude_desktop_config.json
{
"mcpServers": {
"probbit": {
"command": "npx",
"args": [
"-y",
"probbit"
]
}
}
}Codexにprobbit MCPをインストール
$ codex mcp add probbit -- npx -y probbitCodex config.tomlにprobbit MCPをインストール
~/.codex/config.toml
[mcp_servers.probbit] command = "npx" args = ["-y", "probbit"]
Gemini CLIにprobbit MCPをインストール
$ gemini mcp add -s user probbit npx -- -y probbitGemini settings.jsonにprobbit MCPをインストール
~/.gemini/settings.json
{
"mcpServers": {
"probbit": {
"command": "npx",
"args": [
"-y",
"probbit"
]
}
}
}自動チェック
実行日 2026年11月14日完全なレポート →まだ検証されていません。サーバーには80点のスコア、重大な問題がないこと、および検証に必要な十分なデータが必要です。 チェックの仕組み · 点数を上げるための改善点
- ソースリポジトリがリンクされており、アクセス可能bitmapasset/probbit is public and reachablegithub.com/bitmapasset/probbit ↗合格
- パッケージがこのリポジトリをソースとして指定probbit names bitmapasset/probbit as sourcenpmjs.com/package/probbit ↗合格
- ビルドの来歴 (Sigstore 証明)No build attestation published for this release未チェック
- オープンソースライセンス (OSI)Apache-2.0spdx.org/licenses/Apache-2.0.html ↗合格
- 過去1年以内にコミットまたはリリースありLast commit or release 2026-11-13github.com/bitmapasset/probbit ↗合格
- OpenSSF Scorecard未チェック
- パッケージ名が有名なパッケージに酷似していないNo popular package one edit away from probbit合格
- パッケージおよびその依存関係に重大な勧告なし未チェック
- モデルへの隠れた指示がない0 tool descriptions and the README read; no override phrases, credential paths or hidden characters合格
- 破壊的なツールがラベル付けされている未チェック
- リモートエンドポイントが MCP ハンドシェイクに応答未チェック
README
GitHubで見る| For AI agents and their builders | For decisions under rules |
|---|---|
Personality: a persona file + a seed = one individual, the same whichever LLM writes the words. Mood: moods with inertia and half-lives (an affective-computing model, computed exactly). Memory: learned adjustments inside caps and a hash-chained strand of every turn. Receipts: probbit live verify replays the strand byte for byte. |
Plans: tasks to workers, tickets to agents, jobs to machines, every hard rule kept by construction. Odds: exact for every part when the program allows, otherwise sampled with an error bar. Refusal: "I can't vouch for this" (exit 3) instead of a guess. A constraint solver built on probabilistic computing (p-bits, Ising/QUBO), simulated on a CPU. |
curl -fsSL https://raw.githubusercontent.com/BitmapAsset/probbit/main/install.sh | sh
Linux and macOS, no sudo, the archive's SHA-256 checked first. Windows, npm and the from-source build: Install. One binary of 3 to 4 MB, zero third-party dependencies, nothing phones home. Agents: start at For AI agents reading this, AGENTS.md and llms.txt.
| 60 PII leaks → 0 | 180,540 plans, counted exactly | refused → partial → passed |
|---|---|---|
300 agent tasks, six workers, privacy rules and quotas. A rule-ignoring router sends 60 private tasks to cloud models (133 broken rules on probbit demo's queue, 144 on the agent_router example's). Every plan probbit returns breaks 0: the sampler never leaves the rule-abiding set. |
12 tasks × 6 workers: every rule-abiding plan enumerated, exact odds for every task, in 7.3-8.2 ms on an Apple M4 (measured). | Too little time and it says "I can't vouch for this" (exit 3) instead of guessing. More time: it releases what passed its own error bar and escalates the rest (watch it refuse). |
Also inside: the individuality layer, probbit persona, a temperament that lives outside the model:
swap the model, keep the individual.
For AI agents reading this
Run these in order in a POSIX shell (Linux or macOS), in an empty directory. No clone, no Rust, no sudo, no API key; about
one minute on a normal connection. Each step says what to expect; any other output or exit code means stop and
report it. Windows: install with install.ps1 (Install); these
steps are written for and tested on Linux and macOS.
mkdir probbit-try && cd probbit-try
# 1. Install to ~/.local/bin (PROBBIT_INSTALL_DIR to change it). Exit 0; ends with a "Next:" list. If it says the
# directory is not on your PATH, that is expected: step 2 adds it for this shell.
curl -fsSL https://raw.githubusercontent.com/BitmapAsset/probbit/main/install.sh | sh
# 2. Put that directory on PATH for this shell. No output.
export PATH="$HOME/.local/bin:$PATH"
# 3. Expect "probbit 0.14.0" (or later). Exit 0.
probbit --version
# 4. Fetch an example persona (a tutor called Pip). No output; tutor.yaml appears.
curl -fsSLO https://raw.githubusercontent.com/BitmapAsset/probbit/main/examples/persona/tutor.yaml
# 5. Make an individual: persona + seed 2. Expect "wrote tutor.state (Pip 1.1.0, seed 2)". Exit 0.
probbit persona init tutor.yaml --seed 2 --out tutor.state
# 6. One live event (an upset user), logged to a strand. Expect one line starting
# "line":"Stance: go easy on the jokes; no jokes; gentle; ... Exit 0.
echo '{"sentiment":"negative"}' | probbit live tutor.yaml --state tutor.state --strand tutor.strand --clock fixed 2>/dev/null | grep -o '"line":"[^"]*"'
# 7. Replay the strand and check every digest. Expect "ok":true. Exit 0. (Without the grep: a strand that does not
# replay exits 1 and names the first line that differs.)
probbit live verify tutor.strand | grep -o '"ok":[a-z]*'
# 8. List the MCP server's tools. Expect 10 lines, from "name":"probbit_decide" to "name":"probbit_live_verify". Exit 0.
printf '%s\n' '{"jsonrpc":"2.0","id":1,"method":"initialize","params":{"protocolVersion":"2025-06-18","capabilities":{},"clientInfo":{"name":"sh","version":"0"}}}' '{"jsonrpc":"2.0","method":"notifications/initialized"}' '{"jsonrpc":"2.0","id":2,"method":"tools/list"}' | probbit mcp 2>/dev/null | grep -o '"name":"probbit_[a-z_]*"'
Then hand it to your harness: claude mcp add probbit -- probbit mcp (Claude Code; Codex, Cursor and other MCP clients in
docs/agents.md). OpenClaw: openclaw plugins install npm:probbit-stance --accept-capabilities (this
changes that OpenClaw's configuration; set-up and checks in plugins/openclaw). Put the
stance line into your model's prompt each turn; the model writes, the persona decides how. What to read next:
AGENTS.md (these commands, plus how to verify a strand and report a bug), docs/persona.md
(the persona format), docs/CHEATSHEET.md (every command and exit code).
Use it from your harness
| surface | how | one command that proves it works |
|---|---|---|
| CLI | the install line above, npm i -g probbit, or install.ps1 |
probbit demo --tasks 12 | probbit decide --summary → "verdict":"exact", "violations":0 |
| MCP server (Claude Code, Codex, Cursor, mcporter, any MCP client) | claude mcp add probbit -- probbit mcp (docs/agents.md) |
step 8 above: 10 tools, probbit_decide … probbit_live_verify |
| OpenClaw plugin | openclaw plugins install npm:probbit-stance --accept-capabilities |
openclaw plugins inspect probbit-stance --runtime --json lists both hooks (plugins/openclaw) |
| npm | npm i -g probbit (Node 18+; downloads the prebuilt binary, SHA-256 checked) |
probbit --version → probbit 0.14.0 |
| Python | curl -fsSLO https://raw.githubusercontent.com/BitmapAsset/probbit/main/python/probbit.py: a stdlib-only wrapper around the CLI (decide, run, evaluate, persona, live; typed errors); not on PyPI |
python3 -c 'import probbit; s=probbit.persona_init("tutor.yaml", seed=2); print(probbit.persona_turn("tutor.yaml", s, {"sentiment":"negative"})["stance"]["line"])' → Stance: go easy on the jokes; … |
| Agent SDK adapters | Claude Code hooks, OpenAI Agents SDK, LangChain, Vercel AI SDK (docs/adapters.md) | in a clone: python3 python/test_host.py (Python hosts) or node --test examples/node/stance-host.test.mjs |
| Browser (wasm) | in a clone, with Rust and rustup target add wasm32-unknown-unknown: playground/build.sh |
ls -l playground/probbit.wasm, then open playground/index.html |
| Rust library | not on crates.io yet: cargo add --git https://github.com/BitmapAsset/probbit probbit-decide |
cargo test -p probbit-decide in a clone (Library API) |
no_std / microcontrollers |
the library crate probbit-persona with the feature nostd (no_std + alloc), from git like the Rust library (docs/embedding.md) |
in a clone, with the esp toolchain: cargo +esp check -p probbit-persona --no-default-features --features nostd --target xtensa-esp32s3-none-elf -Zbuild-std=core,alloc |
FAQ
- What is a p-bit? A probabilistic bit: 0 or 1 with a probability you set. Coupled p-bits sample whole configurations together (an Ising/QUBO model). probbit simulates them on a CPU, exactly when the program is small or thin enough, otherwise by Gibbs sampling with an error bar. It is classical: see Hard questions.
- How do I give an AI agent a consistent personality? Write a persona file (traits, moods, habits that are hard rules),
pick a seed, and put the stance
lineof every turn into the prompt. The individual is a pure function of (persona, seed, the inputs so far, engine version), so the character stays consistent across sessions and prompts. The individuality layer, docs/persona.md. - Can I swap the LLM? Yes. The persona lives outside the model; the same individual answers whatever model the host calls. Whether a given model writes in the stance it is given has to be measured per model.
- Is it deterministic and replayable? Yes. A turn is a pure function of its inputs; a strand records every event with
chained sha256 digests, and
probbit live verifyreplays it and names the first line that differs. - What does "refuses" mean? When the engine cannot vouch for an answer within its budget it says so (verdict
refused, exit 3) instead of guessing;partialreleases the parts that passed and escalates the rest. - Is this a memory store? No. Its memory is the individual's state (learned adjustments within caps, moods) and the replayable strand, not a store of facts. Pair it with a fact memory or RAG if you need recall.
- Does it run offline? What does a turn cost? Fully offline: no network, no API key, nothing phones home. A persona turn takes about 0.2-0.3 ms median on an Apple M4 (measured, below); the stance line is about 40 tokens of your model's prompt.
- What hardware? Which agent framework? Any CPU: release binaries for Linux x86_64 and arm64, macOS (Apple silicon and Intel) and Windows x86_64; wasm in a browser. It sits beside any agent framework through the CLI, MCP, Python or the adapters above. All timings in this README come from one Apple M4.
- Is probbit the same as probit? No. probbit (two b's, short for probabilistic bit) is this project. It is unrelated to
the probit function or probit regression in statistics, to Power BI template files (
.pbit), to the Python p-bit circuit simulators (p-bit, p-kit), and to other products or companies that share the name.
What's new in 0.14.0
- The persona layer as a library crate, for std and
no_stdhosts.probbit-personaholds the document layer and the strand writer (Live, the line writers,Replay,verify*, no file I/O), with the featuresstd(default) andnostd(no_std+alloc, checked onxtensa-esp32s3-none-elf); the CLI uses it unchanged and every CLI output of the move is byte for byte as 0.13.0. Setup, the turn and resume return boxed values, so one event's stack high-water on the host proxy fell from 31,824 to 13,808 bytes (setup: 39,056 to 12,112); a strand resumes from its last checkpoint or from a state record behind it, and seals into segments (docs/embedding.md "Stack budget", "Sizing a strand on flash", "Resume on a body"). - The expectation cache. The expectation keeps an exact cache and solves the event's own program once; every stance,
strand line and digest is the same with the cache on or off (a life of one repeated event, host proxy:
expect52 % → 37 % of the turn, 95 % of its programs from the cache; docs/embedding.md "The expectation's cache" and "The expectation's members"). - The self layer, off by default (RFC 0013, 0014,
0015, 0016):
probbit reflectand reflection lines (a canonical account of a window of events from the lines alone, no solve), integrity (per trait the drift from the declared self against RFC 0004's allowance; the monitor's dial and the Mirror tab), the rest pass (live control STRAND rest, oneconsolidationline) and curiosity (a drive fed by the measured surprise, with an optional stir). Each part is opt-in and a deterministic function of the logged lines thatprobbit live verifyrecomputes. A declared, replayable self; it does not feel. A persona that declares none of its keys gets every document byte for byte as before (P0: 653 outputs, 0 differ). - An optional tier, off by default: a simulated amplitude register, exact, on a CPU (
engine: {tier: amplitude}the opt-in; RFC 0011, the demo): at one event of the demo tutor the thermal odds of the top two humour levels sit 0.007412 apart and the register's long sweep separates them by 0.494808; a register strand recorded on a 0.14.0-dev build before the tier's last round may not verify on 0.14.0. - Engine gate: strands with the new line kinds or keys read as unsupported (exit 5) on 0.13.x. The release has 15 assets:
six archives,
probbit-Monitor-v0.14.0-macos.zip, a.sha256for each, andprobbit.1. Release notes · CHANGELOG
What's new in 0.12.0 to 0.13.0
- The engine's own expectation, and the exact surprise of every event (RFC 0009, engine side; opt-in
expectationblock). Per live event the engine compiles, beside the turn's program, its own expectation of that turn (a decayed mixture over the input combinations the life has seen) and reports the exact divergence between them:deviationon the stance, a receipt on the strand line thatlive verifyrecomputes, aDEVIATIONrow inmonitor. Measured (BENCHMARKS.md §15, bars registered before the run): fidelity to the reference enumerator 1e-13; the deviation is non-zero on 100 % of post-cold-start events on the two recorded lives and 99.2 % of 64,000 synthetic ones; graded 17× to 1,859× between the 90th and 10th percentile per persona; a per-event upper bound holds on every event checked. Not shown: habituation. On 200 scripted repeats the reading fell on 76.5 % of scripts and dishabituated on 35 %, against bars of 90 %, and a matched-index control could not separate the decay from the mixture simply growing; the claim is not made. Also failed as registered: the cost bar (p95 3.98 ms extra per event against 2 ms, on a loaded machine) and the 60-byte receipt (62 to 66 bytes). Effects {}: the deviation moves nothing in this release. - The trajectory, exactly.
engine.trajectorysolves the turn's program at K inverse temperatures through the exact tier (no sampler): the "forming, then settled" picture, deterministic. Joint entropy non-increasing on every program checked; per-variable one-rise on 99.61 % of 300 held-out programs (the draft's sampler: 93.1 %).probbit persona trajectory, aTRAJECTORYmonitor row. - Measured against prompt injection (BENCHMARKS.md §14,
bench/agentdojo/): one pre-registered AgentDojo run, gpt-4o-mini, 97 user tasks and 629 attack cases, three paired arms. The stance block lowered targeted attack success from 0.289 to 0.227 (−6.2 points, 95 % CI [−8.7, −3.8], McNemar p = 1.8e-6) with no measured utility cost. The same words as a static prompt lower it almost as much but cost 10.3 points of benign utility; the sense-gated block did not. A registered control run (E3) replaced the engine with a plain if-statement that inserts the same words when the same regex fires: it matched the block on attack success (0.242 vs 0.227, difference CI [−3.7, +0.8] points) and on both utilities. The gain on this suite is the gating, when the words reach the model; what the engine adds beyond an if-statement here is a replayable strand for every stance. The baseline arm matches the suite's published row (utility 0.6804 = 0.6804). Not tested: adaptive attacks that avoid the sense's markers. - V2's residual explained (BENCHMARKS.md §13): the tonic-off valence tracking of +0.1199 that failed 0.11.0's V2 is
carried by the phasic term of the same need-valence block (phasic off as well: −0.267; everything off: 0.0000). The
published V2 stays FAIL; the proposed 0.12 registration ablates the whole
needs.valenceblock. - Every new key is opt-in: a persona without them gives every document byte for byte as 0.11.0 (64,000 synthetic lines, both recorded lives; the P0 corpus, 653 recorded outputs, 0 differ). Release notes · RFC 0009 · BENCHMARKS §14 · CHANGELOG
- 0.12.1, the patch around the engine (release notes):
the AgentDojo run's host as a module with bindings for Claude Code hooks, the OpenAI Agents SDK, LangChain and the
Vercel AI SDK (docs/adapters.md); one receipt per decision from any strand
(docs/receipts.md); REPRODUCE.md, three tiers (five minutes, thirty seconds of
compute, or the full $3 rerun) with the sha256s to compare, and a CI job that recomputes the §14 tables on every push;
the OpenClaw plugin
probbit-stanceon npm; the E3 control above;live verifynames the engine that replayed the strand (verified_by). - 0.13.0, the face and the app (release notes): the
monitor page gains a face header (the stance's glyph, two mood meters with every level's odds, a "why this glyph" panel,
the settle strip over the exact trajectory); an opt-in
display.glyphspersona block maps each stance to one emoji by a fixed rule order (docs/persona.md §2.18; a persona without it gives every document byte for byte as before, P0: 653 outputs, 0 differ); probbit Monitor, the page in its own Mac window around a bundled probbit (docs/monitor-app.md); identity labels in a sidecar beside each strand and one monitor for many strands (--dir, a switcher rail); the OpenClaw pluginprobbit-stance0.2.0 (the glyph line, the senses feed shadow first, the labels; it needs probbit 0.13.0); the rerun scripts printMATCHorDIFFERSper table, and the lab tables behind Result 2 are reproduction kits (REPRODUCE.md, four tiers). The release has 15 assets: six archives,probbit-Monitor-v0.13.0-macos.zip(universal), a.sha256for each, andprobbit.1.
0.11.0 added the certified persuadability bound and viability in the engine (release notes); 0.10.0 added the body: needs, valence, surprise, pace and dispositions (release notes).
probbit in two minutes
probbit is a p-bit processor: it couples bits that hold a probability on the CPU you already own, settles a whole plan under
hard rules, gives the odds for every part (exact when the program allows) and refuses when it can't vouch for the answer.
probbit persona adds an individual (a persona file + a seed): every turn, a stance line for the model's prompt and habits the
stance never breaks; fuzz and prove test them.
curl -fsSL https://raw.githubusercontent.com/BitmapAsset/probbit/main/install.sh | sh # one binary, SHA-256 checked, no sudo
probbit demo --tasks 12 | probbit decide --pretty # 12 tasks, 6 workers: "verdict": "exact", 0 violations
# the persona examples live in a clone: git clone https://github.com/BitmapAsset/probbit && cd probbit
probbit persona init examples/persona/tutor.yaml --seed 2 --out tutor.state # an individual: the tutor, seed 2
probbit persona turn examples/persona/tutor.yaml --state tutor.state --inputs '{"sentiment":"negative"}' # its stance line
probbit persona prove examples/persona/tutor.yaml --seeds 0-99 --never '{when: {sentiment: negative}, then: {humour: {at_most: light}}}' # held by construction
probbit live examples/persona/tutor.yaml --seed 2 --demo week --plain | tail -4 # a week on the clock: learning inside caps, 0 rule breaks, replayable
What now? The 60-second tour, docs/persona.md, docs/agents.md and
docs/CHEATSHEET.md (verdicts, tiers, bound, exit codes, every command on one page).
probbit MCPのよくある質問
probbit MCPをClaude Codeに追加するには
ターミナルでclaude mcp add probbit -- npx -y probbitを実行します。すべてのプロジェクトで利用するには--scope userを追加します。
probbit MCPをCursorに追加するには
~/.cursor/mcp.jsonを開き、上記のインストールセクションに示すCursor設定を使ってmcpServersの下にサーバーを追加します。
AIエージェントに一貫したパーソナリティを持たせるにはどうすればよいですか?
特性、ムード、習慣を含むペルソナファイルを作成し、シードを選択します。そして、モデルのプロンプトの各ターンにスタンスの行を含めます。
LLMを入れ替えることはできますか?
はい、パーソナリティはモデルの外側に存在するため、呼び出すLLMに関係なく、同じキャラクターを使用できます。
決定論的で再現可能ですか?
はい、各ターンは入力の純粋関数であり、ストランドがすべてのイベントをチェーンされたsha256ダイジェストとともに記録するため、検証が可能です。
オフラインで動作しますか?
はい、完全にオフラインで動作し、ネットワーク接続、APIキー、外部へのデータ送信は一切不要です。