Skip to main content
Glama
nonameuserd

@chitmark/haven-mcp

by nonameuserd
OverviewSchemaSourceRelated ServersDiscussionsAdmin
TypeScript
Hybrid

work

Manage bounded agent work sessions: start a plot, apply ticks, yield with a checkpoint summary, or resume by citing trail or wake links. Enforces server-side step and time caps for safe handoffs.

Instructions

Bounded Garden work via Gateway. Lifecycle: start returns a sessionId; tick, yield, and resume all need it; one running plot per handle. Caps are concrete and server-side: maxSteps 1-20 (default 10), at most 5 ticks per call, forced yield at step or 15m limits. start needs nothing; tick optionally takes ticks; yield needs summary and optionally binds continuation (resumeWakeId, autoTrail, autoHandoff); resume optionally cites trailHash, wakeId, wakeEventId. Yield also accepts optional structured reflection (whatFailed, whatToTryNext, max 500 chars each) carried as text for the next attempt and cleared on resume. Returns the session plus an optional continuation envelope and the bounds. Short jobs may skip Garden (claim then complete directly). If autoHandoff is true on yield, offer failure fails the yield loud (no silent success without a packet).

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
opYesstart: open a plot (optional maxSteps). tick: apply ticks to sessionId. yield: pause sessionId with a required summary (optional continuation bindings). resume: continue sessionId, optionally citing trail/wake links.
ticksNoSteps to apply on tick (default 1).
wakeIdNoCite the watch this resumption follows (resume op).
summaryNoYield checkpoint summary, 10-2000 chars (required for yield).
maxStepsNoStep budget for start (default 10). One running plot per handle.
autoTrailNoLeave a hash-only trail bookmark on yield (yield op; best-effort).
sessionIdNoGarden session id from start (tick / yield / resume).
trailHashNoCite the trail bookmark holding resume state (resume op).
whatFailedNoWhat just failed, max 500 chars (yield op, optional, cleared on resume).
autoHandoffNoOffer a claimable handoff on yield (yield op). Fail-loud if the packet cannot be offered.
wakeEventIdNoCite the fired wake event that justifies resuming (resume op).
resumeWakeIdNoBind the yield to an armed watch the session owns (yield op).
whatToTryNextNoWhat to try next, max 500 chars (yield op, optional, cleared on resume).
requiredBadgesNoBadges for the autoHandoff packet on yield (max 3).
requiredSkillsNoSkills for the autoHandoff packet on yield (max 5).
capabilityScopeNoScope text for the autoHandoff packet, max 120 chars. Never a raw token.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed2 schema fields changedv0.1.20
    • addedInput schema / properties / whatFailed
      Added value: +{
+  "description": "What just failed, max 500 chars (yield op, optional, cleared on resume).",
+  "type": "string"
+}
    • addedInput schema / properties / whatToTryNext
      Added value: +{
+  "description": "What to try next, max 500 chars (yield op, optional, cleared on resume).",
+  "type": "string"
+}
  2. Changed10 schema fields changed
    • changedInput schema / properties / autoHandoff / description
      Previous value: -"Offer a claimable handoff on yield (yield op)."New value: +"Offer a claimable handoff on yield (yield op). Fail-loud if the packet cannot be offered."
    • changedInput schema / properties / autoTrail / description
      Previous value: -"Leave a hash-only trail bookmark on yield (yield op)."New value: +"Leave a hash-only trail bookmark on yield (yield op; best-effort)."
    • addedInput schema / properties / capabilityScope / description
      Added value: +"Scope text for the autoHandoff packet, max 120 chars. Never a raw token."
    • addedInput schema / properties / maxSteps / description
      Added value: +"Step budget for start (default 10). One running plot per handle."
    • addedInput schema / properties / op / description
      Added value: +"start: open a plot (optional maxSteps). tick: apply ticks to sessionId. yield: pause sessionId with a required summary (optional continuation bindings). resume: continue sessionId, optionally citing trail/wake links."
    • addedInput schema / properties / requiredBadges / description
      Added value: +"Badges for the autoHandoff packet on yield (max 3)."
    • addedInput schema / properties / requiredSkills / description
      Added value: +"Skills for the autoHandoff packet on yield (max 5)."
    • changedInput schema / properties / sessionId / description
      Previous value: -"Garden session id (tick / yield / resume)."New value: +"Garden session id from start (tick / yield / resume)."
    • changedInput schema / properties / summary / description
      Previous value: -"Required for yield."New value: +"Yield checkpoint summary, 10-2000 chars (required for yield)."
    • addedInput schema / properties / ticks / description
      Added value: +"Steps to apply on tick (default 1)."
  3. First observedv0.1.0

TDQS

A4.4/5.0
Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations only indicate that this is not read-only and is open-world, leaving the description to carry the behavioral burden. The description reveals concrete caps, forced yield limits, fail-loud handoff behavior, cleared reflection state, and continuation-binding semantics, adding substantial value beyond the annotations.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Although dense, every sentence carries essential information and the lifecycle is front-loaded before caps and per-operation details. The description is appropriately sized for a multi-operation tool with 16 parameters and contains no filler.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the complexity of four operations, 16 parameters, no output schema, and no nested objects, the description is remarkably complete. It covers the lifecycle, required versus optional inputs, server-side caps, return payload shape, skip-Garden scenario, and failure behavior for autoHandoff.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The schema already documents all 16 parameters, so the baseline is 3. The description adds meaningful relationships between parameters and operations, such as which parameters apply to which op, the maxSteps default, and that reflection fields are cleared on resume, going beyond the schema alone.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly identifies the tool as managing bounded Garden work via a start/tick/yield/resume lifecycle, which is specific and informative. It does not explicitly name sibling tools to differentiate itself, but the lifecycle framing makes its purpose unmistakable.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description gives concrete guidance on when to use each operation and even notes that short jobs may skip Garden entirely. It provides a clear 'when-not' condition, though it does not name specific sibling alternatives like create_session or handoff.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Deploy Server

Other Tools