Repository-native source and plugin
Use this line for the direct GitHub marketplace commands, root source tree, root tests, branded plugin, and repository-native Claude upload.
Evidence: v0.1.5 release checks and custody at their recorded cutoffCannabis cultivation operations · evidence before certainty
CanopyOps structures observations, measurements, calculations, assumptions, facility constraints, and unresolved questions into crop plans, incident workups, harvest reviews, CAPA, runbooks, and handoffs that accountable people can actually inspect.
Advisory decision support for lawful cultivation · Human authority remains explicit · No direct control
Current distribution status
CanopyOps keeps the repository-native source/plugin line separate from the later portable bundle so version numbers cannot borrow evidence from one another.
Use this line for the direct GitHub marketplace commands, root source tree, root tests, branded plugin, and repository-native Claude upload.
Evidence: v0.1.5 release checks and custody at their recorded cutoffUse this line for one self-verifying ZIP with Codex and Claude payloads, detached checksums, a package manifest, and package-specific documentation.
Evidence: static package structure and byte custody, not host activation or field behaviorDo not mix files across versions. Package presence, installation, discovery, invocation, tool health, useful behavior, authority, and verified outcome are separate states.
One crop room, nine different evidence states
CanopyOps keeps state labels attached because a measured value, a calculated derivative, an interpretation, and an approved action do not carry the same authority.
A claim may move forward only when the evidence, method, owner, authority, and verification condition move with it.
Enter through the live cultivation need
Turn facility limits, stage, cultivar information, targets, measurement methods, and open decisions into a reviewable crop plan.
Outputs: facility profile · crop profile · crop plan · room runbookPreserve competing explanations, rank evidence by decision value, and prefer reversible containment while cause remains unresolved.
Outputs: incident workup · evidence gaps · containment · CAPAShow inputs, units, formulas, methods, timing, crop stage, and interpretation for VPD, DLI, irrigation, dryback, runoff, EC, and pH.
Outputs: reproducible calculation · comparison · method caveatSeparate readiness evidence, drying conditions, interpretation, unresolved holds, and the authority still required before release.
Outputs: harvest review · drying log · hold list · release-prep briefAlign records with facility SOPs, current labels, jurisdiction sources, laboratory evidence, named owners, and explicit approval gates.
Outputs: compliance brief · risk register · decision record · shift handoffThe operating kernel
Method, unit, time, sensor location, crop stage, applicable target, and approval determine what a value can legitimately support.
COMPARISON VALUE
A supplied target, remembered range, vendor suggestion, or prior-room value may help compare direction and magnitude.
ACTIVE TARGET
The source, crop stage, location, tolerance, measurement method, and accountable approval are established for the live decision.
Readable operating memory
Primary formats are readable Markdown, CSV, and JSON. Deterministic utilities support calculations, normalization, validation, packaging, and freshness checks when Python is available.
Start with the room as it actually is
Diagnose this cannabis crop incident from my logs and observations. Start with the evidence, preserve competing explanations, and recommend only reversible containment until the cause is supported.
Build a crop plan from this room, cultivar, facility profile, and operating constraints. Separate supplied targets from approved active targets and show the decisions that still need an accountable owner.
Review this batch for harvest readiness. Separate measured evidence, interpretation, unresolved holds, and the approvals required before any release decision.