bioepic-data/ecosim-agent / .agents/skills/plant-trait-target-deriver
plant-trait-target-deriver
VerifiedUse this skill when you need to derive, from web-sourced evidence, typical values for a named plant's annual GPP, LAI, specific LAI or specific leaf area, Rubisco Vcmax at 25oC, Jmax at 25oC, root-to-shoot ratio, rooting depth, leaf protein nitrogen, and leaf chlorophyll nitrogen.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 5.1 KB |
| references/trait_derivation_guide.md | markdown | 3.0 KB |
| references/trait_output_schema.md | markdown | 727 B |
SKILL.md
View on GitHub---
name: plant-trait-target-deriver
description: Use this skill when you need to derive, from web-sourced evidence, typical values for a named plant's annual GPP, LAI, specific LAI or specific leaf area, Rubisco Vcmax at 25oC, Jmax at 25oC, root-to-shoot ratio, rooting depth, leaf protein nitrogen, and leaf chlorophyll nitrogen.
---
# Plant Trait Target Deriver
## Use When
Use this skill when the user gives a plant name such as `Limber Pine`, `maize`, or `switchgrass` and wants a compact trait summary with defensible typical values.
## Target outputs
Derive typical values for:
- annual `GPP`
- `LAI`
- specific `LAI` or specific leaf area
- Rubisco `Vcmax` at 25oC
- `Jmax` at 25oC
- root-to-shoot ratio
- rooting depth
- leaf protein nitrogen concentration
- leaf chlorophyll nitrogen concentration
## Output expectations
Return a compact table or JSON object with:
- `plant_name`
- `accepted_scientific_name`
- `growth_form`
- one row per target trait
- `typical_value`
- `units`
- `provenance`
- `species-sourced`
- `genus-inferred`
- `functional-type-inferred`
- `no-defensible-estimate`
- `rationale`
- `source_links`
## Required behavior
1. Resolve the accepted scientific name first.
2. Determine the growth form:
- tree
- shrub
- grass
- crop grass
- forb/herb
3. Search the web using primary and peer-reviewed sources.
4. Prefer species-level values.
5. If species-level values are unavailable, fall back in order:
- same species in a closely related study system
- same genus
- same functional type
6. Label the provenance explicitly.
7. Never present a guessed value as directly measured.
## Unit conventions
Use these default units unless the user requests otherwise:
- annual `GPP`: `gC m-2 yr-1`
- `LAI`: `m2 leaf m-2 ground`
- specific `LAI` / specific leaf area: `m2 gC-1`
- this is the preferred EcoSIM-facing unit
- if source values use dry-mass units such as `m2 kg-1 DW` or `cm2 g-1 DW`, convert to `m2 gC-1` only when a defensible carbon fraction is available
- if conversion is not defensible, report the source value separately and state that the EcoSIM-unit conversion is unresolved
- `Vcmax25`: `umol m-2 s-1`
- `Jmax25`: `umol m-2 s-1`
- root-to-shoot ratio: `kg root kg-1 shoot` or unitless mass ratio
- rooting depth: `m`
- leaf protein nitrogen concentration: `g N g-1 leaf` or `mg N g-1 leaf`
- leaf chlorophyll nitrogen concentration: `g N g-1 leaf` or equivalent chlorophyll-N fraction
If a source reports nitrogen on an area basis rather than mass basis, preserve the source unit unless a clean conversion is possible from the same source.
## Search strategy
Before doing substantial work, read [references/trait_derivation_guide.md](references/trait_derivation_guide.md).
Suggested sequence:
1. Taxonomy and growth form:
- Kew POWO
- USDA or species-profile sources
2. Productivity and canopy structure:
- flux, forestry, crop, or ecological studies
3. Photosynthetic traits:
- peer-reviewed gas-exchange or trait papers
4. Rooting traits:
- species reviews, crop references, root trait papers
5. Nitrogen allocation traits:
- leaf economics or crop physiology papers
## Trait-specific guidance
- `GPP`
- Prefer site- or ecosystem-specific annual ranges for the target species or strongly species-dominated stands.
- If only NPP or biomass increment is available, do not silently convert to GPP without a defensible method.
- `LAI`
- Prefer field studies, flux-site metadata, crop handbooks, or forest canopy studies.
- specific `LAI` / SLA
- Treat this as specific leaf area unless the user clearly means something else.
- State the interpretation explicitly.
- Prefer the final reported value in `m2 gC-1`.
- `Vcmax25` and `Jmax25`
- Prefer values already standardized to 25oC.
- If a paper reports temperature-response parameters rather than `Vcmax25`/`Jmax25`, do not back-calculate unless the paper provides the needed model clearly.
- root-to-shoot ratio
- Prefer whole-plant or stand-level biomass allocation studies.
- rooting depth
- Prefer effective or typical maximum rooting depth from species or crop references.
- Distinguish between observed depth and model-assumed depth when possible.
- leaf protein nitrogen
- Accept direct protein-N measurements.
- If only total leaf N is available, do not relabel it as protein N unless the source distinguishes the fraction.
- leaf chlorophyll nitrogen
- Prefer explicit chlorophyll-N or nitrogen-allocation measurements.
- If only chlorophyll concentration is reported without nitrogen partitioning, do not convert unless a cited relationship supports it.
## Deliverable format
Prefer a flat table with these columns:
- `trait`
- `value`
- `units`
- `provenance`
- `source`
- `notes`
If the user asks for machine-readable output, emit JSON using the schema in [references/trait_output_schema.md](references/trait_output_schema.md).
## Notes
- This skill is intentionally narrow and should not attempt to derive the entire EcoSIM parameter set.
- Be conservative with `Vcmax25`, `Jmax25`, protein N, and chlorophyll N because these are often study-specific.
- Use exact source links in the final answer whenever web evidence is used.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |