返回技能列表

document-insect-sighting

pjt222
更新于 2 days ago
5 次查看
17
2
17
在 GitHub 上查看
worddesigndata

关于

This skill documents insect observations by capturing location, habitat, photography details, and behavior notes. It helps users perform a preliminary identification and format the data for submission to platforms like iNaturalist. Use it to contribute to citizen science databases or build a personal georeferenced observation journal.

快速安装

Claude Code

推荐
主要方式
npx skills add pjt222/agent-almanac -a claude-code
插件命令备选方式
/plugin add https://github.com/pjt222/agent-almanac
Git 克隆备选方式
git clone https://github.com/pjt222/agent-almanac.git ~/.claude/skills/document-insect-sighting

在 Claude Code 中复制并粘贴此命令以安装该技能

技能文档

Document Insect Sighting

Record insect sightings → structured data + quality photos + citizen sci submission → biodiversity research.

Use When

  • Spot insect → want to doc (personal/research)
  • Contrib to iNaturalist, BugGuide
  • Build systematic journal for habitat/region
  • Support ecological surveys w/ georef photos
  • Beginner learning insect diversity

In

  • Required: Sighting (live in field or recent)
  • Required: Camera/smartphone w/ close-up capability
  • Optional: GPS / phone w/ location
  • Optional: Notebook / field journal
  • Optional: Hand lens (10x) → fine detail
  • Optional: Ruler/coin → scale ref
  • Optional: iNaturalist or equiv account

Do

Step 1: Location + Date + Weather

Capture context before approaching. Many species habitat-specific + seasonally active → metadata as important as photo.

Sighting Record — Context:
+--------------------+------------------------------------------+
| Field              | Record                                   |
+--------------------+------------------------------------------+
| Date               | Full date and time (e.g., 2026-06-15,    |
|                    | 14:30 local time)                        |
+--------------------+------------------------------------------+
| Location           | GPS coordinates if available; otherwise   |
|                    | describe precisely (e.g., "south bank of |
|                    | Elm Creek, 200m east of footbridge")     |
+--------------------+------------------------------------------+
| Elevation          | Meters above sea level if available       |
+--------------------+------------------------------------------+
| Weather            | Temperature (estimate is fine), cloud     |
|                    | cover, wind, recent rain                 |
+--------------------+------------------------------------------+
| Season phase       | Early spring, late spring, summer, early  |
|                    | autumn, late autumn, winter              |
+--------------------+------------------------------------------+

→ Complete context: date + time + precise location (GPS ideal) + weather.

If err: no GPS → describe vs landmarks (trail junctions, buildings, water) → enough to relocate. Weather uncertain → estimate temp range + "overcast"/"clear" over blank.

Step 2: Habitat + Microhabitat

Record where in landscape + immediate substrate/structure.

Habitat Recording:
+--------------------+------------------------------------------+
| Factor             | Record                                   |
+--------------------+------------------------------------------+
| Broad habitat      | Deciduous forest, grassland, wetland,    |
|                    | urban garden, riparian corridor, desert   |
+--------------------+------------------------------------------+
| Microhabitat       | Underside of leaf, bark crevice, flower   |
|                    | head, soil surface, under rock, on water  |
|                    | surface, in flight                       |
+--------------------+------------------------------------------+
| Substrate          | Specific plant species if known, dead     |
|                    | wood, dung, carrion, bare soil, rock     |
+--------------------+------------------------------------------+
| Plant association  | What plant is the insect on or near?     |
|                    | (host plant relationships are diagnostic) |
+--------------------+------------------------------------------+
| Light conditions   | Full sun, partial shade, deep shade       |
+--------------------+------------------------------------------+
| Moisture           | Dry, damp, wet, submerged margin          |
+--------------------+------------------------------------------+

→ Habitat description places insect in ecological context (landscape + microhabitat).

If err: microhabitat hard (insect in flight) → note what flying near / landed on. "In flight, 1m above meadow grasses" over blank.

Step 3: Diagnostic Photos

Photos = most important element. Citizen sci IDs rely almost entirely on image quality.

Photography Protocol:

Shots to take (in priority order):
1. DORSAL (top-down) — shows wing pattern, body shape, coloration
2. LATERAL (side view) — shows leg structure, body profile, antennae
3. FRONTAL (head-on) — shows eyes, mouthparts, antennae base
4. VENTRAL (underside) — if accessible, shows leg joints, abdominal pattern
5. SCALE REFERENCE — place a coin, ruler, or finger near the insect
   for size comparison (do not touch the insect)

Tips for quality macro photographs:
- Get as close as your camera allows while maintaining focus
- Use natural light; avoid flash if possible (causes glare and flattens detail)
- Shoot against a neutral background when feasible (leaf, paper, hand)
- Hold the camera parallel to the insect's body plane for maximum sharpness
- Take multiple shots at each angle — at least 3 per view
- If the insect is moving, use burst mode or continuous shooting
- Photograph the insect in situ first, then closer shots if it remains
- Include at least one photo showing the insect in its habitat context
- If wings are open, photograph quickly — the pattern may change when
  wings close (especially butterflies and dragonflies)

→ ≥3 usable photos: dorsal + lateral + scale. Ideal ≥5 across angles.

If err: insect moves before all angles → prioritize dorsal (top-down) → most diagnostic for ID. Sharp dorsal > multiple blurry. Flies before any photo → sketch body shape + colors from memory immediately.

Step 4: Behavior + Interactions

Behavioral obs add ecological value photos can't capture.

Behavioral Notes:
+--------------------+------------------------------------------+
| Category           | Record what you observe                  |
+--------------------+------------------------------------------+
| Activity           | Feeding, flying, resting, mating,        |
|                    | ovipositing (egg-laying), burrowing,     |
|                    | grooming, basking                        |
+--------------------+------------------------------------------+
| Movement           | Crawling, hovering, darting, undulating   |
|                    | flight, walking on water, jumping        |
+--------------------+------------------------------------------+
| Feeding            | What is it eating? Nectar, pollen, leaf   |
|                    | tissue, other insects, dung, sap?        |
+--------------------+------------------------------------------+
| Interactions       | Other insects nearby? Being predated?     |
|                    | Ants attending? Parasites visible?        |
+--------------------+------------------------------------------+
| Sound              | Buzzing, clicking, stridulation (wing or  |
|                    | leg rubbing)? Silent?                    |
+--------------------+------------------------------------------+
| Abundance          | Solitary individual, a few, many (swarm,  |
|                    | aggregation)?                            |
+--------------------+------------------------------------------+
| Duration           | How long did you observe?                 |
+--------------------+------------------------------------------+

→ ≥3 behavioral obs: activity + movement + abundance.

If err: brief encounter (lands, flies) → record what you did see + duration. "Resting on leaf, solitary, flew when approached, 5 sec" = useful.

Step 5: Preliminary ID → Order

No need for species. Placing in order narrows ID + helps reviewers.

Quick Key to Major Insect Orders:

1. Count the legs.
   - 6 legs → insect (proceed below)
   - 8 legs → arachnid (spider, tick, mite) — not an insect
   - More than 8 legs → myriapod (centipede, millipede) — not an insect
   - Wings but hard to count legs → likely insect; look at wings

2. Examine the wings.
   - Hard front wings (elytra) covering body → Coleoptera (beetles)
   - Scaly wings, often colorful → Lepidoptera (butterflies/moths)
   - Two wings + knob-like halteres → Diptera (flies)
   - Four membranous wings + narrow waist → Hymenoptera (bees/wasps/ants)
   - Half-leathery, half-membranous front wings → Hemiptera (true bugs)
   - Large, transparent wings + long abdomen → Odonata (dragonflies/damselflies)
   - Straight, narrow, leathery front wings → Orthoptera (grasshoppers/crickets)
   - No wings, laterally flattened, jumps → Siphonaptera (fleas)
   - No wings, pale body, in wood or soil → Isoptera (termites)

3. If unsure, note: "Order uncertain — resembles [description]"

→ Preliminary ID to order (e.g., "Coleoptera — beetle") or honest "order uncertain" + description.

If err: no clear match → record body shape + wing type + leg count. iNaturalist accepts "Insecta" as start. Honest "unknown" > forced guess.

Step 6: Submit to Citizen Sci Platform

Upload → experts + community verify/refine.

Submission Checklist for iNaturalist (or equivalent):

1. Upload photographs — start with the best dorsal shot
2. Set location — use the map pin or enter GPS coordinates
3. Set date and time of observation
4. Add initial identification (order or family if known; "Insecta" if not)
5. Add observation notes:
   - Habitat and microhabitat
   - Behavior observed
   - Approximate size
   - Any sounds produced
6. Mark as "wild" (not captive/cultivated)
7. Set location accuracy — use the uncertainty circle to reflect GPS precision
8. Submit and monitor for community identifications

Data Quality Tips:
- Observations with 3+ photos from different angles get identified faster
- Including habitat context in one photo helps remote identifiers
- Adding a size reference dramatically improves identification accuracy
- Responding to identifier questions speeds up the process
- "Research Grade" status requires 2+ agreeing identifications at species level

→ Complete obs submitted w/ photos + location + date + prelim ID → community review.

If err: no net in field → save photos + notes locally, upload later. Most platforms allow backdated. No account → store in journal, value for learning + upload later.

Check

  • Date + time + precise location → before approach
  • Weather + habitat documented
  • ≥3 photos from diff angles
  • ≥1 photo w/ scale ref
  • Behavior + activity noted
  • Prelim order ID attempted (or honestly marked unknown)
  • Submitted to citizen sci or structured journal

Traps

  • Approach too fast: Many flee rapid approach. Move slow + no shadow over subject. Photo from far, then close.
  • Ignore habitat context: Insect on white wall → loses ecology. Always ≥1 in-situ photo.
  • Single photo: One img insufficient. Wings/legs/antennae only from specific angles.
  • Forget scale: No size ref → 5mm beetle = 50mm beetle in photo. Coin/ruler/finger for scale.
  • Force ID: Confident-but-wrong ID creates noise for researchers. "Insecta" / "order unknown" always OK, preferred over wrong genus/species.
  • Skip negatives: "No insects on milkweed patch" = valuable absence data. Record what checked, not just what found.

  • identify-insect — detailed morphological ID beyond order
  • observe-insect-behavior — structured ethological protocols
  • collect-preserve-specimens — when physical specimen needed
  • survey-insect-population — scaling sightings → systematic pop surveys

GitHub 仓库

pjt222/agent-almanac
路径: i18n/caveman-ultra/skills/document-insect-sighting
0
agentsagentskillsai-assisted-developmentclaude-codeskillsteams

相关推荐技能

content-collections

Content Collections 是一个 TypeScript 优先的构建工具,可将本地 Markdown/MDX 文件转换为类型安全的数据集合。它专为构建博客、文档站和内容密集型 Vite+React 应用而设计,提供基于 Zod 的自动模式验证。该工具涵盖从 Vite 插件配置、MDX 编译到生产环境部署的完整工作流。

查看技能

polymarket

这个Claude Skill为开发者提供完整的Polymarket预测市场开发支持,涵盖API调用、交易执行和市场数据分析。关键特性包括实时WebSocket数据流,可监控实时交易、订单和市场动态。开发者可用它构建预测市场应用、实施交易策略并集成实时市场预测功能。

查看技能

creating-opencode-plugins

该Skill帮助开发者创建OpenCode插件,用于接入命令、文件、LSP等25+种事件。它提供了插件结构、事件API规范和JavaScript/TypeScript实现模式,适合需要拦截操作、扩展功能或自定义事件处理的场景。开发者可通过它快速构建响应式模块来增强OpenCode AI助手的能力。

查看技能

sglang

SGLang是一个专为LLM设计的高性能推理框架,特别适用于需要结构化输出的场景。它通过RadixAttention前缀缓存技术,在处理JSON、正则表达式、工具调用等具有重复前缀的复杂工作流时,能实现极速生成。如果你正在构建智能体或多轮对话系统,并追求远超vLLM的推理性能,SGLang是理想选择。

查看技能