본문으로 건너뛰기
cccskills
무료GitHub에 공개

check-feature-support

Check whether a CARLA feature is usable in the current environment and has a verified procedure. Review known failures, inspect build settings, and identify differences between client, server, and source versions before planning work.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

원문원문 설명 보기

Answers "can CARLA do X here, and is there a vetted procedure for it" before you promise anything, and detects client/server/tree version skew — a support matrix over the whole collection (covered by a skill / works but undocumented / present but untested / broken or removed), an index of the features that are verified BROKEN on 0.10.0 (gbuffers crash the server, map layers do nothing, Landmark.waypoint is None, standalone asset packages are gone), and a probe that reports what this machine's build flags and running server actually offer. Use when the user asks whether CARLA supports something, when a feature has no skill, when a UE4 tutorial's steps do not exist, or before improvising a procedure.

설치 방법 보기

이럴 때 유용합니다

  • Check whether a CARLA feature has a verified procedure
  • Review known failures before following a UE4 tutorial
  • Inspect build flags for ROS 2 or RSS support
  • Investigate client and server version differences

포함된 파일(5)

  • SKILL.md9.8 KB
  • references/matrix.md7.6 KB
  • scripts/check_env.sh2.4 KB
  • scripts/env.sh2.2 KB
  • scripts/support.sh17.5 KB

SKILL.md(원문)

설치하기 전에 에이전트에게 전달되는 지시 내용을 확인할 수 있습니다.

Is this feature usable, and is there a procedure for it?

Paths. scripts/… and references/… below are relative to the directory holding this SKILL.md. Your working directory is the user's project, not that directory, so prefix them with its absolute path or the command is not found.

The failure this prevents is specific: an agent asked to use an uncovered CARLA feature reconstructs a plausible procedure from UE4-era documentation, and it is wrong. Every row in this skill's [broken] list used to be a working, well documented UE4 workflow. Reciting those steps on ue58-dev produces a silent no-op, an empty result, or a crashed server.

So answer from this skill instead of from memory, and when the answer is "no vetted procedure exists", say that and offer to investigate.

Instructions

Progress:
- [ ] Step 1: bash scripts/check_env.sh   (no blockers — it reports probe depth)
- [ ] Step 2: matrix — find the feature's status
- [ ] Step 3: broken — read this BEFORE promising anything on 0.10.0
- [ ] Step 4: probe — confirm against this machine's build and server
- [ ] Step 5: version — only if a measurement looks impossible (skew check)
source scripts/env.sh

bash scripts/support.sh matrix     # the table (works with no CARLA)
bash scripts/support.sh broken     # verified-broken index, with evidence
bash scripts/support.sh probe      # this machine: flags, tree, live server
bash scripts/support.sh version    # the identity tuple + skew warnings

version: am I reasoning about the thing I am talking to?

"Which CARLA is this" has five answers that can disagree — installed client, server, source tree, content branch, engine — and a disagreement is silently wrong rather than loudly wrong. version prints all five plus the build flags, and warns on the two skews that cause wasted debugging:

  • client ≠ server. CARLA usually still connects, then misbehaves in ways that look like your own bug.
  • installed client older than the tree HEAD, or than a wheel built from it. You read a source change that the client you are calling does not contain. This is not hypothetical: it happened while these skills were being written, and it is why one API was briefly reported as removed when it had only moved class.

Note what version deliberately does not claim: that matching versions mean matching behaviour. ENABLE_ROS2, ENABLE_RSS and ENABLE_OSM2ODR are compile-time, and the map-layer breakage is a property of the content branch — two servers reporting 0.10.0 can differ completely. Use matrix for behaviour; use version to establish that your measurements are even about the right target.

Engine lines, and the version number

UE 5.8 (ue58-dev) is the line that continues, and it is the one released as CARLA 1.0. UE 5.5 (ue5-dev) is an earlier revision of the same line, not a fork: PythonAPI/carla/src/Sensor.cpp is byte-identical between them. So the verdicts below apply to both engines unless a row says otherwise, and the five features 5.5 lacks are catalogued in [[check-ue5-limitations]].

On the version string. Every measurement here was taken on a pre-release ue58 build, which reports 0.10.0 — that is the number version will print on those, and the number 5.5 reports today. Read 0.10.0 below as naming the UE5 line: 1.0 renamed it, it did not change the behaviour underneath. Where a verdict is genuinely engine-specific the row says so.

The four verdicts, and what each obliges you to do

VerdictMeaningWhat to do
[skill]a vetted procedure exists hereinvoke that skill; do not improvise
[works]verified working, no skilluse the API directly, say it is unvetted
[untested]in the build, never exercised heresay so, offer to investigate, do not write steps
[broken]verified broken or removed on 0.10.0say so, name the alternative if one exists

What probe adds that the matrix cannot

Two features are gated by compile-time flags that no runtime check reveals, and every -D option must be repeated on each re-configure — so the CMake cache is the only honest source:

  • ENABLE_RSS — sensor.other.rss appears in the blueprint library either way, and cannot function without it. On the tree measured here it is OFF.
  • ENABLE_OSM2ODR — the OSM→OpenDRIVE half of digital twins. OFF here.
  • ENABLE_ROS2 — no native ROS 2 interface at all without it, and it cannot be switched on at runtime ([[run-autoware-ue58]], [[build-carla-ue58]]).

probe also reports tree-visible facts that decide whether a documented procedure can work at all: whether Co-Simulation/ exists (it does not on ue58-dev), whether RecastBuilder sits in Util/DockerUtils/dist (it does not, which is why map-import navmesh generation silently produces nothing), whether the Autoware integration is shipped, and whether the multi-GPU sources are there.

Examples

Example 1: "grab the GBuffer normals from the camera"

broken first. On 0.10.0 listen_to_gbuffer() asserts on Stream.has_value() and takes the server down, while the whole API surface still exists. Tell the user it crashes the simulator; offer sensor.camera.normals instead, which is a real sensor.

Example 2: "hide the buildings so I can see the road network"

matrix shows map layers as [broken]. enable_environment_objects is the mechanism that works — that is [[toggle-env-objects]], a [skill] row.

Example 3: "set up SUMO co-simulation"

[broken]: ue58-dev has no Co-Simulation/ directory at all, while UE4 ships Sumo, PTV-Vissim, Chrono and Carsim bridges. Do not port the UE4 instructions blind — say the tooling is absent and ask whether they want it investigated.

Example 4: "can I use the RSS sensor?"

probe. The blueprint exists; if ENABLE_RSS=OFF the honest answer is "not in this build — it needs a rebuild with -DENABLE_RSS=ON, and nobody here has exercised it afterwards".

Example 5: "add a vehicle to ue58"

[untested] — to be done; there is no skill for it in this version. The pipeline is known and gets as far as a vehicle that registers, spawns and reports four wheels in get_physics_control(); it does not move, because the physics asset does not survive an editor restart on the skeletal mesh. Deferred to a later version rather than shipped half-working.

So: say it is not supported yet, offer to investigate, and do not adapt [[import-carla-ue58-walker]]'s procedure. Walkers bind to a shared skeleton; vehicles need a per-vehicle skeleton plus a persisted physics asset, which is the one step still unsolved.

Troubleshooting

Problem: the matrix disagrees with the CARLA docs Cause: the public docs describe the UE4 tree. Solution: trust probe and the evidence in references/matrix.md — they were taken from the tree in front of you. Docs are the weaker source here.

Problem: probe reports nothing about build flags Cause: no root variable set, or the tree is unconfigured, or CARLA_PRESET names a preset with no cache. Solution: export CARLA_UE58_ROOT (or CARLA_UE4_ROOT), and set CARLA_PRESET to a preset you have actually configured.

Problem: a feature is missing from the matrix Cause: the matrix covers what has been examined; CARLA is larger than that. Solution: say it is unexamined — which is not the same as unsupported — and offer to check the tree. Do not extrapolate from a neighbouring row.

Problem: a measurement contradicts the source you just read Cause: version skew — most often an installed client older than the tree. Solution: version. On the machine these skills were written on it caught a wheel built one day after the installed client, which is exactly how a moved API gets misreported as a removed one.

Problem: this skill is stale after an upgrade Cause: every verdict is stamped to ue58-dev HEAD 718efd7cc / 0.10.0. Solution: re-run the measurements in references/matrix.md; each row records how it was established so it can be re-checked cheaply.

Outputs

Nothing is written and nothing is started: all four modes are read-only. matrix and broken print text; probe prints what this machine's variables, CMake cache, tree layout and running server report; version prints the identity tuple and exits non-zero for nothing — read its MISMATCH lines, they are advisory but they invalidate measurements.

Per-feature evidence — file paths, line numbers, the measurements behind each [broken] row, and what a future skill would have to verify to promote a row — is in references/matrix.md.

리뷰

아직 리뷰가 없습니다. 사용해 본 소감을 남겨 주세요.

댓글

로그인 없이 쓸 수 있습니다. 사용 후기, 질문, 작동하지 않은 환경 등을 편하게 남겨 주세요.

아직 댓글이 없습니다. 첫 댓글을 남겨 보세요.

보내면 바로 공개됩니다. 이메일 주소나 전화번호 같은 개인 정보는 쓰지 마세요. URL은 2개까지입니다.

0 / 1,000자

이 양식은 reCAPTCHA로 보호되며 Google 개인정보처리방침 및 서비스 약관이 적용됩니다.

같은 저장소의 스킬

Add a message type to CARLA's native ROS 2 interface. Create its C++ data structure, serialization code for transmission, type identity hash, registration, and tests, with Docker-based hash calculation and build checks.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

  • Adding a standard ROS 2 message type missing from CARLA
  • Defining a custom carla_msgs message
  • Investigating type-hash warnings
carla-simulator/carla1.5만2026년 10월 20일 업데이트

Extend CARLA’s native ROS 2 interface to publish sensor data or receive vehicle commands. Covers topic implementation, registration, data connections, and verification, plus adding middleware, the communication layer used to exchange messages.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

  • Publish obstacle sensor data to ROS 2
  • Add a topic for vehicle speed commands
  • Integrate another communication middleware
carla-simulator/carla1.5만2026년 10월 20일 업데이트

Summarize ScenarioRunner results to see which scenarios passed and which checks failed. Compare runs before and after changes, or analyze recorded driving to produce measurements and plots of vehicle distance and deviation from the lane center.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

  • Checking scenario pass and fail results
  • Comparing results before and after controller changes
  • Plotting vehicle distances from a recorded run
carla-simulator/carla1.5만2026년 10월 20일 업데이트

Checks a local CARLA checkout for missing API documentation, outdated references, broken links, and pages absent from navigation. Produces a grouped Markdown report without building the project or running the simulator.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

  • Reviewing documentation before a release
  • Finding missing or outdated API references
  • Checking broken links and navigation gaps
carla-simulator/carla1.5만2026년 10월 20일 업데이트

Read and display CARLA object bounding boxes, then project them onto a camera image as 2D rectangles. Export an annotated PNG and box coordinates in JSON for object detection annotation work.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

  • Adding vehicle boxes to a camera image
  • Exporting box coordinates for detection annotations
  • Checking vehicle dimensions
carla-simulator/carla1.5만2026년 10월 20일 업데이트

Builds CARLA's ue4-dev branch on Linux, producing the Unreal Engine 4.26 editor, server modules, and an installed Python client package. Fetches required content, optionally compiles native ROS 2 support, and verifies the build against a source-built server.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

  • Building CARLA ue4-dev from source
  • Compiling CARLA with native ROS 2 support
  • Building the Python client for an existing environment
carla-simulator/carla1.5만2026년 10월 20일 업데이트

carla-simulator의 스킬 모두 보기

이 스킬의 문제 신고