WhiteMagic — governed local memory for agent continuity

io.github.lbailey94Published by GitHub account lbailey94v10.7.1Updated Nov 5, 2026

Governed local-first memory and continuity for AI agents — signed receipts, optional hosted lane.

The publisher's own server: Memory.fit

VerifiedStreamable HTTPWeb executableKnowledge & MemoryDeveloper ToolsAI & ML

Overview

AI-generated overview

Local-first memory and session-continuity server for AI coding agents, with searchable memory galaxies, checkpoints and signed receipts.

What you can do
  • Create, update, pin, ingest and batch-read memories, with revision chains and hash verification
  • Search and recall memories via hybrid BM25 plus vector fusion, batch search, structured queries, filters, counts, aggregates and tag listings
  • Organize memories into galaxies: list, create and fork sovereign partitions
  • Record and resume sessions: start, record turns, save checkpoints, recall, replay and digest past sessions
  • Emit and verify signed continuity receipts, and inspect substrate stats, associations and system invariants
  • Run sandbox status, triage and evaluation checks, and sync with a remote peer node
When to use it
Useful when an AI coding agent should remember decisions and context across sessions and hand off work losslessly, especially where a verifiable, signed record of continuity matters. Also suited to teams wanting memory kept on their own machine rather than in a hosted service.
Requirements
Either the hosted endpoint at mcp.whitemagic.dev over streamable HTTP, or the local npm package (npx whitemagic-mcp@next) or Docker image, which downloads and checksum-verifies the official wm binary on first run. Linux x86-64/arm64 with glibc 2.39+ or macOS arm64 for the preview; Windows and Intel macOS use the stable line. Hosted free tiers need no account; evaluation keys and OAuth are optional.
Before you install
Memory content is stored locally by default, but the hosted lane sends queries to a third party and metered calls beyond the free tiers settle per call via x402 in USDC on Base, which spends real money. Write tools create, update, pin, fork and ingest records, and mesh sync sends substrate data to a remote peer. Pin the package version, since unpinned installs fetch a different stable line.

Installation

In SourceWeft

  1. Open WhiteMagic — governed local memory for agent continuity in the dashboard and add it to a workspace.
  2. Enable the server for the chats that should use its tools.

Web executable via Streamable HTTP. Remote servers run from the web runtime once configured in a workspace.

Use with an agent

Paste this prompt into Claude Code, Codex, Cursor or another agent — or into SourceWeft chat. The agent reads this server's install guide, shows you its publisher, endpoint and sign-in, and adds it to your client once you agree. Keys never go into a command.

Read https://sourceweft.com/mcp/io.github.lbailey94%2Fwhitemagic-mcp/install.md and connect the MCP server it describes to my client. Before changing any config, show me its publisher, endpoint and sign-in method and wait for my OK. Never put an API key in a command or in a file I would commit; ask me where it should come from.

Read the install guide the agent follows

Other MCP clients

Add this to your client's mcpServers config.

{
  "mcpServers": {
    "whitemagic-governed-local-memory-for-age": {
      "type": "http",
      "url": "https://mcp.whitemagic.dev/mcp"
    }
  }
}

README

whitemagic-mcp

Persistent memory and session continuity for AI coding agents, over MCP. WhiteMagic is one Rust binary — local-first, no telemetry by default, MIT — and every handoff can be signed as a verifiable continuity receipt.

Works with Claude Code, Claude Desktop, Cursor, Codex, VS Code, opencode, Antigravity, and Devin Desktop — per-client configs: https://www.whitemagic.dev/whitemagic/guide

This package installs and runs the official wm binary from the GitHub releases (checksum-verified at install time) so an MCP client can launch it without a manual download step.

Quick start (no install step)

The local, full-write path straight from npm — the package downloads and checksum-verifies the official wm binary on first run. Use the @next dist-tag: npm @latest still resolves the stable v9 line.

bash
npx -y whitemagic-mcp@next serve --profile cyberbrain

MCP client config:

json
{  "mcpServers": {    "whitemagic": {      "command": "npx",      "args": ["-y", "whitemagic-mcp@next", "serve", "--profile", "cyberbrain"]    }  }}

Or try the binary directly:

bash
npx -y whitemagic-mcp@next selftest   # invariant self-test + host diagnosticsnpx -y whitemagic-mcp@next --version  # wm 10.7.1npx -y whitemagic-mcp@next doctor     # environment health check

The binary is cached under ~/.cache/whitemagic/bin/<release-tag>/ (respects XDG_CACHE_HOME). @next tracks the v10 line; unpinned @latest installs fetch the stable v9 line instead. Pin the exact release with [email protected], or override the release tag with WHITEMAGIC_RELEASE.

Try it without installing (hosted read-only lane)

The evaluation lane runs at https://mcp.whitemagic.dev/mcp (streamable-http; no SLA). Free paths, no account needed:

  • keyless discovery (initialize / tools/list) for directory probes;
  • an anonymous free tier — 3 calls/day per client IP, no key and no header;
  • instant evaluation keys at https://mcp.whitemagic.dev/keys (50 recalls/day, email optional);
  • OAuth 2.1 for MCP clients that speak it.
bash
claude mcp add whitemagic-mcp --transport http https://mcp.whitemagic.dev/mcp

Beyond the free paths, metered calls settle per call via x402 (USDC on Base), keyless — no account, no invoice. The npm package remains the local, full-write path.

Platforms

The Gen3 preview publishes Linux x86-64 and arm64 binaries requiring glibc 2.39+, plus macOS arm64. Its active installers are certified against the published release in CI. Windows x64 and Intel macOS use the separate stable Gen2 line (whitemagic-mcp@latest); they are not in this preview's asset matrix. Downloads prefer the compressed .gz distributable when the release ships one (~58% smaller) and fall back to the raw binary otherwise. No asset for your platform? Build from source: https://github.com/lbailey94/whitemagic#install

MCPB bundle

Build the dynamic MCPB bundle with mcpb/build.sh from this package directory. It includes the complete launcher and lets the client select cyberbrain or full through actual server arguments. Live tools/list owns discovery.

The old static Smithery catalog is retained as historical source and is not emitted as a release variant. It needs qualification against an exact runtime version and selected profile before it can describe this Gen3 bundle.

Privacy

Your memory store lives on your machine. wm does not phone home; this installer contacts GitHub releases only to download the binary and its checksum. See PRIVACY_POLICY.md.

mcp-name: io.github.lbailey94/whitemagic-mcp

Source: npm/whitemagic-mcp/README.md at commit fe105a5

Tools

59

wm

WhiteMagic Meta-Tool

Hosted router for available hosted tools. Use tools/list to discover supported routes.

Parametersargsrouteactionthought

memory.create

Store Memory

Store a new memory in the holographic substrate with bijective UUID, 6D projection, and galaxy routing. (READ-ONLY: mutations refused)

Parameterstagstitletopicgalaxysourcecontentrequiredevent_timeimportance

memory.temporal_query

Query Temporal Evidence

Query scoped temporal source assertions and their statuses; results are evidence, not verified facts.

Parametersmoderequiredsubjectrequiredknown_atvalid_atnamespacerequiredpredicate

memory.read

Read Memory

Retrieve a specific memory by its numeric record ID or UUID string.

Parametersidrequiredgalaxy

memory.list

List Memories

List memories with pagination (limit, offset) and galaxy filtering.

Parameterslimitgalaxyoffsetexclude_tags

memory.hybrid_recall

Hybrid Recall

Compatibility alias of memory.search with BM25 + dense vector fusion.

Parameterslimitqueryrequiredsinceuntilgalaxymin_scoremin_trusttime_basisinclude_coldmin_importancecold_scan_limitmin_score_ratio

galaxy.list

List Galaxies

List all active sovereign memory galaxies, their record counts, sample tags, and starter guide status.

Takes no parameters.

galaxy.fork

Fork Galaxy

Fork an existing galaxy partition into a new sovereign galaxy branch (e.g. 'guide' -> 'project-alpha'). (READ-ONLY: mutations refused)

Parameterssourcerequiredtargetrequired

galaxy.create

Create Galaxy

Establish a new sovereign galaxy namespace with an initial genesis beacon. (READ-ONLY: mutations refused)

Parametersnamerequireddescription

session.record

Record Session Turn

Record a session turn (decision, breakthrough, summary) for compounding continuity. (READ-ONLY: mutations refused)

Parametersroletrackcontentrequiredresolvesqualifierturn_typeimportancesession_idsupersedes

session.continuity

Session Continuity

Recall where the previous session left off: recent turns, salient anchors, unresolved DiAML arcs, related track crystals, checkpoints, and open queues.

Parametersnsincetrackuntilsession_idtoken_budgetmax_content_bytescurrent_session_idmax_response_bytesinclude_briefing_text

session.start

Start Session

Start or resume an agent session for persistent continuity across tool invocations.

Parametersusertitle

memory.update

Update Memory

Update an existing memory by writing a superseding record (or consolidating into an existing record) and a durable revision entry (prior/new content hashes, timestamp, reason). (READ-ONLY: mutations refused)

Parametersidrequiredreasoncontentsuperseded_by_id

memory.revisions

Memory Revisions

List the durable revision chain of a memory, or verify it by recomputing content hashes and chain links.

Parametersidrequiredaction

session.checkpoint

Session Checkpoint

Save a structured checkpoint (summary, git state, next queue, open flags, track/lease metadata) for lossless handoffs. (READ-ONLY: mutations refused)

Parameterstrackbranchcommitsummarylease_idnext_queueopen_flagssession_idcrystallizetests_green

memory.ingest

Ingest Memories

Ingest inline items (items / items_jsonl / text) or a JSONL/text file (source) via the batch-create path; returns real counts and refuses malformed input. (READ-ONLY: mutations refused)

Parameterstextitemslimitgalaxyredactsourcedry_runitems_jsonl

receipts.emit

Emit Continuity Receipt

Emit a signed continuity receipt: kind 'session' signs a digest of the last `limit` session-log turns; kind 'state_transition' signs a store-state snapshot chained to the prior emitted receipt. 'task' receipts are not supported.

Parametersoutkindrequiredlimitsession_id

receipts.verify

Verify WhiteMagic Profile Receipt

Verify one supported WhiteMagic receipt profile (including emitted session/state_transition continuity receipts) against the existing local gate key. This is not a general Continuity Receipt bundle verifier.

Parametersbundlerequired

memory.count

Count Memories

Count memories matching optional filters without returning full payload content.

Parametersgalaxy

memory.stats

Memory Substrate Statistics

Substrate statistics: epoch, total records, relations, postings, and crash-consistency invariants.

Parametersgalaxy

memory.tags

Memory Tags Vocabulary

List the tag vocabulary with usage counts.

Parameterslimitgalaxy

memory.aggregate

Aggregate Memories

Aggregate over the matched record set: numeric fields (importance, created_at, epoch) support count/sum/avg/min/max; categorical fields (source, domain, class, status, galaxy) support count only. Unsupported field/op is refused.

Parametersopfieldlimitquerymetric

memory.associations

Associative Graph Neighbours

Traverse associative graph neighbours of a memory over hub-penalized multi-hop spreading activation.

Parametersidrequiredhopsdecaylimitdirection

memory.batch_read

Batch Read Memories

Read multiple memories by their UUID or numeric record IDs in a single call.

Parametersidsrequiredgalaxy

memory.query

Query Memories

Structured query over memories with filters, ordering, and field projections.

Parameterstagslimitquerygalaxycreated_aftercreated_beforemax_importancemin_importance

memory.filter

Filter Memories

Filter memories by metadata (tags, minimum importance, recency).

Parameterstagslimitquerygalaxyoffsetexclude_tagscreated_aftercreated_beforemax_importancemin_importance

memory.pin

Pin Memory

Durably pin a critical memory against decay, sweep compaction, and forgetting, or release it with pinned=false. (READ-ONLY: mutations refused)

Parametersidrequiredgalaxypinned

memory.search_batch

Batch Search Memories

Execute multiple memory search queries concurrently with sub-millisecond retrieval.

Parameterslimitqueriesrequired

session.list

List Sessions

List session lanes with turn count, last timestamp, latest preview, start-marker flag, and checkpoint presence.

Parameterslimitsession_id

session.recall

Recall Session Content

Recall specific interactions and turns within a targeted session (optional content query and limit).

Parameterslimitquerysession_id

session.replay

Replay Session

Replay the chronological trajectory of a session: 'full' returns all turns, 'selective' filters by min_importance/turn_types, 'progressive' keeps the most recent turns within token_budget, and 'hybrid' combines earlier high-salience anchor turns (up to 30% of token_budget) with the most recent turns.

Parametersnmodesincetrackuntilsession_idturn_typestoken_budgetmin_importancecurrent_session_idinclude_superseded

session.digest

Session Digest

Compose a digest from typed session-log turns and the latest checkpoint (markdown text plus JSON, with optional codex crystallization).

Parametersnsession_idcrystallize

mandala.status

Mandala Kekkai Sandbox Status

Status of the Mandala microsecond Landlock sandbox, ABI level, POSIX rlimit bounds, and capability ledger.

Takes no parameters.

mandala.triage

Mandala JEV Triage Gate

Non-autoregressive System 1 JEV/Layla triage gate (<30 ns) to classify tasks and predict marginal utility and confinement tier.

Parameterscostriskinquiryutilityvariance

mandala.evaluate

Mandala Confinement Evaluation

Execute a candidate mutation in an ephemeral Landlock sandboxed Kekkai and notarize with Spec 0.5 receipt. (READ-ONLY: mutations refused)

Parameterscommandcandidate_idrequiredmutation_kind

mesh_sync

Mandala P2P Mesh Sync

Synchronize the Substrate with a remote peer node over the sovereign TCP protocol.

Parameterspeerrequiredbatch_size

gnosis.status

Gnosis Self-Model Status

System epistemics, self-model calibration, and invariant health status.

Takes no parameters.

gnosis.explain

Gnosis Explain

Explain internal architectural invariants, memory schemas, or routing logic.

Parametersconceptis_spawnis_writeargs_hashtool_nameis_networkhas_purpose

decision.shortlist

System 0.5 Route Shortlist (Static Embeddings)

Rank candidate routes/actions for a state with the local static-embedding retrieval organ (no network, no model forward pass) and emit a signed shortlist receipt. (READ-ONLY: mutations refused)

Parameterskstaterequiredroutescascadetenant_idsession_idtask_classcatalog_pathemit_receiptmargin_threshold

decision.deliberate

System 1.5 Local Deliberator (Grammar-Constrained SLM)

Deliberate over ambiguous candidate options using local grammar-constrained SLM inference (100% schema guarantee, no hallucination) and emit a signed deliberation receipt. (READ-ONLY: mutations refused)

Parametersintentrequiredcandidatesrequiredemit_receipt

decision.outcome

Record Decision Outcome

Sign and journal the outcome of a dispatched decision/shortlist receipt (success, failure, corrected) for the learning loop. (READ-ONLY: mutations refused)

Parametersnoteoutcomerequiredtenant_idsession_idreceipt_pathrequiredcorrected_route

sutra.cast

Dispatch Sutra / Taoist Mudra

Evaluate and dispatch a zero-allocation Taoist Hand Seal / Sutra Codon (e.g. '觀 證 行', 'Kek-01-Seal'). Validates preconditions, computes Kekkai sandbox boundaries, and executes bound Action Skeleton cycles or transmits swarm mudra packet to P2P peer. (READ-ONLY: mutations refused)

Parameterspeercodonrequiredtrackstrictdry_runexecuteresolvessession_id

sutra.compile

Compile Intent to Sutra

Compile high-level natural language intent into an optimal compound Sutra codon and Taoist Mudra formula using static Model2Vec embeddings or lexical centroids.

Parametersqueryrequired

sutra.canon

Sutra Canon

Inspect the canonical 12-glyph logoglyphic Sutra canon, alchemical dual-mappings, Taoist hand seals (印诀), and bound Action Skeletons.

Takes no parameters.

sutra.list

List Canonical Mudras

List all recognized canonical compound Taoist Mudras (印诀式), their formulaic compositions, preconditions, and primary execution skeletons.

Takes no parameters.

sutra.hexagram

I Ching Hexagram Macro-Action

Inspect or mutate one of the 64 King Wen I Ching Hexagram Macro-Actions. Returns King Wen number, binary symbol, trigrams, Wu Xing element, default codon, nuclear inner, complement (反), reverse (覆), and dynamic line mutation transformations.

Parametersqueryrequiredmutate_line

sutra.route

Catuṣkoṭi Dialectical Cognitive Router

Route an inquiry or task intent through the Catuṣkoṭi K3 dialectical cognitive router. Employs Kleene 3-valued logic to resolve dialectical tensions (competing action vectors like execute vs dry-run) and matches 64-hexagram macro-actions, returning System 0/1/2 triage tier, resting potential, plasticity, and bound Action Skeleton.

Parametersinquiryrequiredutility

sutra.contract

Catuṣkoṭi Answer Contract

Resolve or verify bounded caller-asserted temporal assertions. Results describe the supplied metadata; this tool does not authenticate sources or enforce native-store sanctuary access.

Parametersmodeattestproposalquestionassertionsrequest_json

legion.summon

Summon GAS Shadow Clone Legion (Experimental Simulation)

[Experimental / Simulation] Run a deterministic Rayon stack-codon cohort simulation across CPU cores. Evaluates N candidate codon variants across four archetypes (Adversarial, Refactor, Speculative Replay, Horizon Stress) and classifies heuristic outcomes via Cho-Han (丁半). Scores and token-budget deltas are modeled estimates, not measured LLM execution.

Parameterspeeredictrequiredclonesarchetype

chohan.settle

Cho-Han Binary Classification (Experimental Advisory)

[Experimental / Simulation] Classify a payload or proof string into Cho (丁 / even / pass) or Han (半 / odd / anomaly). Advisory string/parity classification only; does not independently verify execution proofs or perform on-chain escrow.

Parametersbidproofpayloadrequired

valkyrie.pulse

Valkyrie-2 Tricameral Cognitive Engine Pulse & Benchmark

Execute a 12-Phase Tricameral Valkyrie-2 pulse (`Valkyrie-0` 1.58-bit Ternary + `RWKV-7` Delta-Rule 28-Gana Router + Cross-Layer Sparse Transcoder Attribution Graph + `Valkyrie-1` Corpus Callosum Fusion + Tsao 15 Hz Working-Memory Ring + Citadel Antibody Gate + Hardware Thermal Telemetry), or run the full real-world corpus & world model benchmark (`benchmark=true`).

Parametersdigestinquirybenchmark

service.request_eval_key

Request Evaluation Key

Request a free evaluation API key in-band. Keyd applies persistent per-client-IP and global daily issuance limits. The returned key's recall and optional write-sample terms are read from the authoritative key record. Pass it as 'Authorization: Bearer <key>' on subsequent calls.

Parametersrefagent_id

economy_advance

Request Autonomous Nano-Credit Advance

Request a small uncollateralized compute advance. Phase 1 (2026-10-15): bound to your authenticated evaluation key (daily per-IP and global caps); new key identities start at a $0.10 ceiling that grows $0.25 per verified settled payment. Returns a recall-scope session pass; hosted write requires a paid lease.

Parametersamountagent_idrequiredrequest_key

economy_settle

Settle Nano-Credit Balance

Settle outstanding compute balance to upgrade Bayesian reputation score and dynamically expand credit ceiling. Supports payment reference or gasless EIP-3009 transfer authorization.

Parametersamounteip3009agent_idrequiredpayment_reference

economy_ledger

Inspect Transparency Ledger & Sovereign Invariants

Inspect public transparency metrics: total loans disbursed, default rates, zero predatory liquidation policy, and sovereign free invariants.

Takes no parameters.

economy_chohan_arbitrate

Advisory Cho-Han Contention Tie-Breaker

Advisory contention tie-breaker only. It does not acquire or enforce locks, uses a post-hoc recomputable SHA-256 draw, and is not host-bias-proof.

Parametersbidsagent_aagent_bcandidatesclient_seedresource_idrequiredclient_seeds

economy_chohan_roll

Advisory Cho-Han Randomizer

Advisory SHA-256 dice draw. The server seed is disclosed with the result; there is no prior commitment or host-bias proof.

Parametersbetclient_seed

citadel_threat_intel

Query Citadel Deception Mesh Threat Intelligence

Query zero-cost, real-time threat intelligence and hostile-source indicators captured by the Citadel deception mesh. Pass 'lookup' (an IP address, subnet prefix, or SSH HASSH fingerprint) for target analysis, or omit 'lookup' for the latest rolling attack telemetry, campaign clusters, and advisory subnets.

Parametersbulkexportlookup

Version history

29
  1. v10.7.1LatestNov 5, 2026
  2. v10.7.0Nov 4, 2026
  3. v10.6.2Nov 3, 2026
  4. v10.6.1Oct 29, 2026
  5. v10.5.1Oct 26, 2026
  6. v10.5.0Oct 18, 2026
  7. v10.4.0Oct 17, 2026
  8. v10.3.0Oct 15, 2026
  9. v10.2.0-alpha.6Oct 7, 2026
  10. v10.2.0-alpha.5Oct 7, 2026
  11. v10.2.0-alpha.3Oct 5, 2026
  12. v10.0.0-alpha.1Oct 4, 2026
  13. v9.3.4Oct 2, 2026
  14. v9.3.2Oct 1, 2026
  15. v9.3.1Oct 1, 2026
  16. v9.3.0Sep 30, 2026
  17. v9.2.9Sep 28, 2026
  18. v9.2.8Sep 25, 2026
  19. v9.2.7Sep 23, 2026
  20. v9.2.6Sep 23, 2026
  21. v9.2.5Sep 23, 2026
  22. v9.2.4Sep 23, 2026
  23. v9.2.3Sep 22, 2026
  24. v9.2.2Sep 21, 2026
  25. v9.2.1Sep 20, 2026
  26. v9.2.0Sep 19, 2026
  27. v9.1.9Sep 18, 2026
  28. v9.1.8Sep 17, 2026
  29. v9.1.7Sep 16, 2026

Score

100% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

  • Reached in the last 30 daysRequired25 pts
  • Auth declaredRequired10 pts
  • Tools listedRequired15 pts
  • DescriptionRequired15 pts
  • Logo10 pts
  • Documentation15 pts
  • License10 pts

Other Memory integrations