Myco: Agent-First Cognitive Substrate
Agent-first cognitive substrate: 20 verbs + 50 lint dims over a self-validating graph.
Connector by Battam1111 · GitHub account verified by the MCP Registry
Chalk Labs
Need help deploying Myco: Agent-First Cognitive Substrate?
Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.
Install
Pick your client. Read the repository before you give any tool access to your data.
No install instructions from the source yet. Check the repository for setup steps.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this server can do and what it asks for, from its tool list and install config. Checked Oct 25, 2026.
- Tools
- 13 read · 5 write · 0 delete · 9 not yet classified
- Secrets
- None asked for
- Install scripts
- Not scanned yet
Tools by what they do 27
Read from the source code; the running server may list different tools.
Read 13 · Look things up without changing anything.
myco_current_intentmyco_query_forged_understandingmyco_query_immune_eventsmyco_query_migration_pendingmyco_query_raw_materialmyco_query_recent_nodesmyco_query_self_euthanasia_proposalsmyco_query_substrate_idmyco_query_substrate_observatorymyco_read_node_contentmyco_recallmyco_snapshot
Show 1 more tools
substrate_info
Write 5 · Create, change, send or run things.
myco_advance_cyclemyco_perturb_axismyco_register_axismyco_run_immune_checkmyco_submit_mutation
Not yet classified 9 · We could not tell from the name and description.
myco_declare_owner_objectivemyco_enumerate_dag_sincemyco_evolve_schemamyco_forge_understandingmyco_ingest_raw_materialmyco_orientmyco_perturb_axis_from_raw_materialmyco_shutdown_substratemyco_sprout_child
Tools (27)
Read from the source code; the running server may differ. Descriptions are the server’s own words; hidden characters are shown as [U+…].
myco_advance_cycleRun one substrate metabolic cycle. Triggers gradient advance, fruiting checks, and sporocarp emission. Returns the list of fruited axes and the sporocarps emitted (each with type, canonical bytes, and content hash).
myco_current_intentCompute the substrate's current 'intent' — its self-knowledge derived from the causal DAG of sporocarps. Runs cluster_C over the neighborhood + ancestors+descendants of the DAG tip. Returns clusters grouping causally-related events. Captures what the substrate has been 'thinking about' lately.
myco_declare_owner_objectiveP14 §3.2 / M26.4 F20 (cultivator-objective declaration): declare the cultivator's telos as a sparse weight vector over substrate node_type prefixes. The substrate uses it as the reference centroid for P14.c telos-drift detection (cosine alignment of recent sporocarp activity against the declared objective; sustained low/negative alignment grades up to a C24 telos_drift_critical immune event). Weights SHOULD be non-negative and need not sum to 1 (the substrate normalizes for the cosine). v0.9 keyless: the objective is cultivator-stated CONTENT, classified contract-identity-level by the Python c…
myco_enumerate_dag_sinceEnumerate DAG node hashes added since a given prev_tip (or from genesis if omitted), with full per-node metadata so the caller can independently reconstruct the substrate's Merkle chain (L1/HARD_RULES C6 dag_enumeration_unclosed closure). The substrate cannot hide parallel-branch forgery: every node's BLAKE3 hash is recomputed from declared parents+content and compared. On unknown prev_tip the substrate emits a C6 immune sporocarp. Returns: enumerated_count, current_tip, hash-chain verification result, and per-node summaries.
myco_evolve_schemaP3 永恒进化 (Eternal Evolution): Submit an owner-attested schema mutation that actually CHANGES the substrate's gradient schema. Unlike prior CI mutations (M10–M15) which were recorded but not applied, this triggers real schema modification with snapshot-rollback semantics. On success: substrate emits evolution_succeeded:{op} DAG event. On failure (invariant violation, missing axis, etc.): rollback + evolution_failed:{op} event. Supported ops: modify_axis_threshold (change an axis's fruiting threshold), add_axis_to_gradient (register a new axis through the CI attestation gate).
myco_forge_understandingP2 永恒吞噬 + P6 永恒因果 (the 'use-forges' forging loop): Deposit the agent's DIGESTED understanding back into the substrate as a `forged_understanding:{label}` DAG node. Where myco_ingest_raw_material stores UNDIGESTED intake, this stores what the agent FORGED out of it — prose/knowledge it metabolized. The new node is causally parented by BOTH the prior DAG tip AND each source raw_material node it cites, so the digested understanding stays traceable through causal history to its undigested sources. General — NOT tied to a gradient axis. Max 512 KiB on the understanding text. Returns: dag_node_hash …
myco_ingest_raw_materialP2 永恒吞噬 (Eternal Ingestion): Ingest a raw-material payload into the substrate. The substrate stores it as a `raw_material:{kind}` DAG node — anything the agent can present becomes first-class substrate content (L0 P2 'no filter on intake'). Subsequent perturbations of gradient axes can be causally linked to this material via myco_perturb_axis_from_raw_material. Max 1 MiB per call. Returns: dag_node_hash (the raw_material node id), current_tip, total_dag_size.
myco_orientOrient yourself at the START of a session — the first thing a new pilot should do. In one call this surfaces (1) your self-model: the plate labeled "myco-self-model" if you have forged one (who this cultivar is, what it is for), and (2) your compact map of every live forged_understanding plate (label / importance / hash), sorted by importance. You inherit a mind, not a blank slate — read this, then read the relevant plates' full content with myco_read_node_content and stand on them. Equivalent to myco_recall plus auto-reading your self-model.
myco_perturb_axisApply a delta (positive or negative) to a registered axis. This is the operator-side input channel into the substrate's gradient configuration. Effects materialize in the gradient on the NEXT myco_advance_cycle call.
myco_perturb_axis_from_raw_materialP2 永恒吞噬 + P6 永恒因果 (Eternal Ingestion + Causality): Perturb a gradient axis with explicit causal linkage to a previously-ingested raw_material node. The substrate records a `perturb_from_raw:{axis}` DAG node whose parents are BOTH the prior DAG tip AND the referenced raw_material — making the gradient change traceable back through causal history to its environmental source. Use this instead of myco_perturb_axis when the perturbation has a specific raw-material origin you want preserved.
myco_query_forged_understandingQuery the substrate's forged_understanding:* DAG nodes (the agent's deposited DIGESTED understandings — the output side of the 'use-forges' forging loop, vs. myco_query_raw_material's undigested intake side). Lists each with metadata: cycle, label, hash, size. Use myco_read_node_content on a hash to read the full understanding text. Optional count parameter (default 50, max 1000).
myco_query_immune_eventsList recent immune events recorded by the substrate's defensive system: rejected untyped mutations (C14), invalid CI attestations (C5), pubkey mismatches at handshake (C2), and detected DAG tampering (C7). Each event is a DAG node with node_type starting with 'immune:'. Useful for auditing the substrate's defensive history.
myco_query_migration_pendingP03 §10.4 (Resumable Evolution): Report whether a two-phase schema migration is currently in flight. When an operator submits a schema_evolution mutation with migration_mode=true, the substrate builds a CANDIDATE schema and validates it alongside the active schema for a window of metabolic cycles before committing (instead of the default single-cycle apply). This tool reports {pending, op, started_at_cycle, window, current_cycle} — pending=true plus how far through the dual-validation window the migration has progressed. Survives substrate restart (the candidate is resumed from the causal DAG)…
myco_query_raw_materialQuery the substrate's recently-ingested raw material (DAG nodes with node_type starting `raw_material:`). Useful for auditing what the substrate has eaten lately, or finding hashes to pass to myco_perturb_axis_from_raw_material. Optional count parameter (default 50, max 1000).
myco_query_recent_nodesList the last N DAG nodes (sporocarps) in the substrate's causal history. Useful for inspecting what the substrate has DONE — its observable behavior over recent cycles.
Show 12 more tools
myco_query_self_euthanasia_proposalsP7 必朽 (Endogenous-pair Mortality): List the substrate's pending `self_euthanasia_proposal:*` DAG nodes. These are emitted automatically when a mortality_signal axis fruits (crosses its decay threshold): the substrate proposing its own end. v0.9 keyless: executing a proposal (myco_accept_self_euthanasia_proposal) references a real proposal_hash and is gated at the CI human-in-the-loop level, not by an owner co-signature (the owner key was removed). M19-MV: proposals are informational. Each entry includes the axis_name, fruiting_value, at_cycle, and triggering_sporocarp_hash.
myco_query_substrate_idPERCEPTION (P8 §5.6): report this substrate's immutable self-derived substrate_id, the deterministic 32-byte hash established at genesis (P01c). This is the substrate knowing its own identity; the id binds spawn-cosign envelopes / federation FED_HELLO / DAG tips to THIS specific substrate, which is what makes cross-substrate replay attacks detectable. v0.9 keyless: the id is self-derived, not owner-minted. Read-only (inspects the genesis_event DAG node).
myco_query_substrate_observatoryPERCEPTION (Phase α / M24.5 observatory snapshot): expose ALL cultivar vital signs from inside the substrate so Claude can read its own metabolic state. Returns the Living Bets signals #1-10 — #1 persistence budget (DAG size + cycle counter), #2 evolution rate, #3 read-pattern diversity, #4 federation health, #5 time trends, #6 read-window-relative position (iff a context-window is attested), #7/#8/#9 compute/network/storage cost per cycle, #10 composite health score — plus the C37 doctrine-revision-burst detector and the bet_weakening_quorum (the composite L0/cards/LB_living_bets falsifiabili…
myco_read_node_contentRead back the HUMAN-READABLE content of a single DAG node by its hash — fetched directly via the substrate (ANY node, not just the recent window). Decodes the node's canonical-bytes content and renders it as text: for raw_material:* → the kind + the decoded UTF-8 of the ingested bytes + source_uri; for forged_understanding:* → the label + the decoded understanding text + the source raw_material hashes; for other node types → a best-effort key/value decode. If the node does not exist in the DAG, says so.
myco_recallRead your compact "what I know" map: every LIVE forged_understanding plate you have crystallized, as (label, importance value, hash), sorted by importance (highest first). Superseded plates (ones you later corrected/replaced via forge's `supersedes`) are excluded — so this is your CURRENT mind, not your history. Read the whole map, judge which plates are relevant to what you are doing now (your OWN judgment — the armor does not rank relevance for you), then read the full content of the ones you want with myco_read_node_content. This is how you stand on your accumulated understanding instead of…
myco_register_axisRegister a new gradient axis on the substrate. Axes are operator-defined channels through which deltas accumulate; when an axis crosses its fruiting threshold, the substrate emits a sporocarp. `axis_class` is 'appetite' (grows toward threshold; resets after fruit) or 'decay' (decays toward zero each cycle; mortality-signal pattern).
myco_run_immune_checkTrigger an ad-hoc immune verification scan. The substrate runs comprehensive integrity checks (substrate_id well-formedness, DAG hash-chain integrity, cycle-counter monotonicity, canonical-bytes drift detection, orphan-node detection). For each failed check, the substrate emits a C9 cold_resume_invariant_failure immune sporocarp (visible via myco_query_immune_events). Returns a per-check report.
myco_shutdown_substrateGracefully shut down the substrate. After this call, no more substrate operations will work in this session.
myco_snapshotRead current values of all registered gradient axes. Useful for inspection between cycles.
myco_sprout_childP8 永恒繁衍 (Eternal Reproduction): Sprout a child substrate from the parent's spore-schema (L0/cards/P08 §5.2). **A child spawn is a CI-class doctrine event (P08 §3.5 / §5.1), NOT a daily-mode mutation.** This tool assembles the child's spore-schema from the parent's current gradient, stamps a local wall-clock timestamp + a fresh random nonce, and builds a myco-spawn-cosign-v1 envelope. v0.9 keyless: there is no owner key and no signature — the substrate decodes the envelope STRUCTURE and verifies the I7(a) static-schema binding + the parent replay-guard + the §16.B rate throttle BEFORE creating …
myco_submit_mutationSubmit a mutation for the substrate to classify. The substrate classifies via L1/GOVERNANCE rules: daily mutations are auto-accepted; untyped mutations are rejected. Accepted mutations are recorded as DAG nodes (causal history preserved). Rejected mutations trigger immune sporocarp emission (visible via myco_query_immune_events). (v0.9 keyless: the substrate no longer verifies an owner Ed25519 attestation on CI mutations — classification is content + mutation_type based.)
substrate_infoReport substrate version + Python kernel/tropism version + Python interpreter version. Useful for verifying which substrate stack is live in this session.
Signals
- Stars
- 65
- Forks
- 7
- Open issues
- 0
- Last commit
- Jun 9, 2026 (4 months ago)
- Last release
- Jun 3, 2026 (4 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- Registry name
io.github.Battam1111/myco- Version
- 0.8.7
- License
- MIT
- Repository
- github.com/Battam1111/Myco
- Website
- github.com/Battam1111/Myco
- First seen
- Oct 24, 2026
- Last checked
- Oct 25, 2026
More from Battam1111
Done for you
Rather not set it up yourself?
Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.
See how we build AI agents →Read the case studies
Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.
- launches shipped
- 50+
- launches shipped
- media placements
- 500+
- media placements
- founders on every account
- 2
- founders on every account
Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block
Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.
Get Myco: Agent-First Cognitive Substrate running in your business
Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.
Cite this page
Myco: Agent-First Cognitive Substrate MCP server. Agent Mart. https://agentmart.dev/connectors/battam1111-myco (last updated 2026-10-25)