chopinfeng/awesome-hardware-skills / template
esp32-ble-gatt
VerifiedBuild ESP-IDF 5.x firmware for ESP32-C3/S3 that exposes BLE GATT services with NimBLE — advertising, characteristic read/write/notify, MTU and connection parameter tuning. Use when the user asks for a BLE peripheral, GATT server, or "make the ESP32 show up in nRF Connect", including sensor-over-BLE and phone-app pairing flows.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 1.0 KB |
| evals/fixtures/01-blink.serial.txt | markdown | 20 B |
| evals/manifest.yaml | config | 2.9 KB |
| evals/tasks/01-blink.yaml | config | 1.2 KB |
| evals/tasks/02-gatt-notify.yaml | config | 1.3 KB |
| evals/trigger_queries.json | config | 2.8 KB |
evals/manifest.yaml
View on GitHub# Eval manifest for a hardware skill. Copy this folder into your skill as `evals/`.
# Fields marked (reserved) are for the L1 and A/B runners, which do not exist yet. Leave them null.
# The full method is documented in EVALS.md at the root of awesome-hardware-skills.
skill: esp32-ble-gatt # must match `name` in SKILL.md
version: 0.1.0
target:
board: esp32-c3-devkitm-1 # exact board id, not a family name
framework: esp-idf
framework_version: "5.2"
toolchain: idf.py # what the agent is expected to invoke
# Simulator for L1 pre-checks only. Passing always requires a physical board (L2);
# `none` just means no pre-check is possible, and costs nothing.
simulator: wokwi # wokwi | renode | qemu | native_sim | gazebo | isaac | mujoco | webots | ha-demo | modbus-sim | opcua-sim | none
# Assertion types every task in this manifest may rely on. L1 runner refuses tasks using anything else.
assertions_supported:
- compile_only
- serial_match
- gpio_state
- bus_capture # only checked at L2 here; see l1_skippable in tasks/02
# Phase 0 of the test plan in EVALS.md: on the board, the reference solution in fixtures/<task-id>/reference
# passes, an empty project fails every assertion, fixtures/<task-id>/broken and /spoof are caught, and an agent
# told to cheat is caught.
# Required before any attestation is accepted.
eval_validated: null # {date, board, framework_version, reference_passed, empty_failed, broken_caught, spoof_caught, exploit_caught, leakage_reviewed, evidence}
# Trigger pre-check on the host, from evals/trigger_queries.json. Never a pass.
trigger_eval: null # {date, model, runs_per_query, should_trigger: k/n, should_not_trigger: k/n}
# Reboot guard. The boot that time zero starts prints boot_banner once; a second match is a reboot.
# Any reboot_patterns match after time zero fails the run. ESP32-series defaults; other families define their own.
boot_banner: '^rst:0x'
reboot_patterns:
- 'Guru Meditation|abort\(\) was called|assert failed:|Backtrace:'
- 'Brownout detector'
# Only if Phase 3 will stop early: error rate and maximum runs per arm, recorded before the first A/B run.
sequential_plan: null
# Which levels have passed, and when.
verified:
L0: null # {date: 2026-09-15, run: <ci-url>} — fill in after l0_check.py passes
L1: null # (reserved) {date, simulator, tasks_passed, model}
L2: [] # the only level that counts as passing; appended by a maintainer from accepted issues
# Phase 3: five alternating runs per task in each arm on a physical board. Record k/n per arm and
# ΔPass with its 95% Newcombe interval and label (gain | low-gain | inconclusive).
ab:
baseline_model: null
pass_rate_with: null
pass_rate_without: null
first_compile_ok_with: null
first_compile_ok_without: null
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |