Weavebench Cua Reproduce
AMAP-ML/LongHorizon-Harness
Reproduce CUA-Harness experiments on WeaveBench from a GitHub checkout.
The reports room, where every report people file on GitHub (Report a Bug and Send Feedback from the app, and issues opened by hand) is followed from arrival to the release that ships its fix, in a…
$ npx skills add pdcgomes/redlamp --skill redlamp-reports -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install pdcgomes/redlamp redlamp-reports --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/pdcgomes/redlamp.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.cursor/skills/redlamp-reports .claude/skills/redlamp-reports && rm -rf skills-srcUse ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.
Claude Code skills documentation · loads skills from .claude/skills/
Install the "redlamp-reports" agent skill from https://github.com/pdcgomes/redlamp/tree/main/.cursor/skills/redlamp-reports into .claude/skills/redlamp-reports/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "redlamp-reports", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/pdcgomes/redlamp/tree/main/.cursor/skills/redlamp-reportsType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add pdcgomes/redlamp --skill redlamp-reports -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install pdcgomes/redlamp redlamp-reports --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/pdcgomes/redlamp.git skills-src && mkdir -p .agents/skills && cp -r skills-src/.cursor/skills/redlamp-reports .agents/skills/redlamp-reports && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "redlamp-reports" agent skill from https://github.com/pdcgomes/redlamp/tree/main/.cursor/skills/redlamp-reports into .agents/skills/redlamp-reports/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "redlamp-reports", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add pdcgomes/redlamp --skill redlamp-reports -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install pdcgomes/redlamp redlamp-reports --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/pdcgomes/redlamp.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/.cursor/skills/redlamp-reports .cursor/skills/redlamp-reports && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "redlamp-reports" agent skill from https://github.com/pdcgomes/redlamp/tree/main/.cursor/skills/redlamp-reports into .cursor/skills/redlamp-reports/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "redlamp-reports", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/pdcgomes/redlamp.git --path .cursor/skills/redlamp-reports--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add pdcgomes/redlamp --skill redlamp-reports -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install pdcgomes/redlamp redlamp-reports --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/pdcgomes/redlamp.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/.cursor/skills/redlamp-reports .gemini/skills/redlamp-reports && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "redlamp-reports" agent skill from https://github.com/pdcgomes/redlamp/tree/main/.cursor/skills/redlamp-reports into .gemini/skills/redlamp-reports/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "redlamp-reports", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install pdcgomes/redlamp redlamp-reportsInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add pdcgomes/redlamp --skill redlamp-reports -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/pdcgomes/redlamp.git skills-src && mkdir -p .github/skills && cp -r skills-src/.cursor/skills/redlamp-reports .github/skills/redlamp-reports && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "redlamp-reports" agent skill from https://github.com/pdcgomes/redlamp/tree/main/.cursor/skills/redlamp-reports into .github/skills/redlamp-reports/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "redlamp-reports", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add pdcgomes/redlamp --skill redlamp-reports -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install pdcgomes/redlamp redlamp-reports --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/pdcgomes/redlamp.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/.cursor/skills/redlamp-reports .opencode/skills/redlamp-reports && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "redlamp-reports" agent skill from https://github.com/pdcgomes/redlamp/tree/main/.cursor/skills/redlamp-reports into .opencode/skills/redlamp-reports/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "redlamp-reports", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
redlamp-reportsThe reports room, where every report people file on GitHub (Report a Bug and Send Feedback from the app, and issues opened by hand) is followed from arrival to the release that ships its fix, in a…
Redlamp Reports is an agent skill from pdcgomes/redlamp. The reports room, where every report people file on GitHub (Report a Bug and Send Feedback from the app, and issues opened by hand) is followed from arrival to the release that ships its fix, in a canvas the owner triages from. A bug he marks Fix it gets its own agent chat that reproduces it, fixes it, pushes to main and replies on the issue with the cause, the fix and the release; a suggestion he accepts goes onto the roadmap, one he rejects is closed with his reason. Use when a bug report or feedback comes in…
Its SKILL.md is about 2.7k tokens, which your agent loads only when the skill is triggered. The skill folder holds 3 other files (for example `bug-agent.md` and `room.py`).
It sits in Testing & QA, covering QA and bug reports. It works with GitHub. The repository describes itself as: A native, open-source RAW photo editor for Mac, iPad and iPhone that feels immediately familiar to Lightroom users. The licence is MPL-2.0.
5 steps, taken from the first numbered list in SKILL.md.
Read from SKILL.md and the folder at commit bc52683. It shows what the files ask for, not the result of running them.
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
Ships script files (Python), which the agent can run.
Shell commands in SKILL.md call:
ghpython3From the folder's file list and the shell code blocks in SKILL.md.
No URLs in SKILL.md. Its commands use gh, which can reach the network depending on how they are called.
From URLs in SKILL.md, links to its own repository left out.
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Redlamp Reports loads about 2.7k tokens when it runs. Until then it costs about 182 tokens; SKILL.md has 1,600 words of instructions outside code blocks.
Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.
The automated check found no risky patterns in SKILL.md.
Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.
The full file from pdcgomes/redlamp at commit bc52683, republished under its MPL-2.0 licence (© pdcgomes). 1,600 words, ~2,738 tokens.
.claude/skills/redlamp-reports/SKILL.md (or your agent's skills folder). This skill also uses 3 other files; get the full folder from GitHub.Every report is followed in one canvas, the reports room: the reporter's own words, the owner's triage, each bug's agent and how far it has got, the fixes on main and the release each is expected in, and the suggestions with the owner's decision. The owner triages everything by hand for now; nothing is fixed, answered, closed or added to the roadmap until he says so on the board.
| What | Where |
|---|---|
| The room | ~/.cursor/projects/Users-pedrogomes-src-darkroom/canvases/reports-room.canvas.tsx, from template.tsx. room.py sync writes its DATA block; edit only its room object (decisions and links). A change to the rendering goes into the template and the room together. |
| The owner's clicks | reports-room.canvas.data.json beside it, under triage: { "<n>": { "decision": "fix" | "close" | "accept" | "reject" | "answer", "at": "…", "reason"?: "…", "token"?: "…" } }. Only the canvas writes it. |
| The room's notes | ~/.cursor/projects/Users-pedrogomes-src-darkroom/reports-room/: reports/<n>.json per report (what agents record), log.jsonl, thumbs/. Written only through room.py, so agents working at once never write the same file. |
| The script | room.py: status, sync, watch, claim, note, log (its docstring has every option). Run the main checkout's copy, from anywhere. |
| The bug agent's brief | bug-agent.md, which Fix it hands each bug's agent |
| How reports are filed | packages/RedlampUI/Sources/Feedback (the sheet and Your Reports), web/lib/feedback.ts (the relay), .cursor/rules/tracker-issues.mdc (Issues people file) |
Read ~/.cursor/skills-cursor/canvas/SKILL.md once per session before the first edit of the room. Link the room in every reply that changes it: [Reports room](/Users/pedrogomes/.cursor/projects/Users-pedrogomes-src-darkroom/canvases/reports-room.canvas.tsx).
python3 .cursor/skills/redlamp-reports/room.py status. Never take states from memory. It lists the reports waiting for triage, the decisions waiting to be applied, the agents at work, and the closed ones, with each one's commits on main and branches, the release it's expected or was released in, and the replies. It folds in the owner's clicks.room.py note <n> --read "…".docs/research/research-tracker.md, its item text and phase), follows an existing row, or a duplicate of one: room.py note <n> --proposal '{"kind": "new row", "id": "TON-31", "text": "…", "phase": "P3"}'.room.py sync, and say plainly what you didn't check.The owner decides on the Triage tab. Each card shows the report's own sections (What happened, What I expected, Steps to reproduce; What I'd like to do; Message), its area, how often, the version and Mac, the replies, a thumbnail of the first screenshot, a read and a proposal when an agent has written them.
fix with a token and opens a new chat for that bug alone, prompted to follow bug-agent.md. The agent claims the report with the token, which ties the chat to the board.accept; its label says what the proposal does.reject or close with the owner's reason, in a sentence written for the reporter.answer and opens a chat that answers it (Questions, below).A Fix it that no agent has claimed shows under "Fix it, no agent has picked it up yet", with Start its agent again.
When a report is reopened after its fix (the reporter says it isn't fixed, or asks for more), it goes back to Triage. Its card says who reopened it and when, and shows the earlier round: the commits, the release they shipped in, its reply, and Open the first agent's chat. Clicks, reads and fixes from before the reopen no longer count; room.py finds the reopen in the issue's events and keeps the earlier round in the report's history (rounds). Fix it again opens a new agent chat whose prompt names the earlier fix and the first agent's chat, and asks it to find out why that fix wasn't enough before changing anything. When the agent claims the report, the earlier round is filed under rounds in its note.
Take each recorded decision in turn, then record it with room.py note <n> --triage <decision> --applied "<what you did>" (and --tracked <ID> where it became one), so the board stops showing it as waiting.
gh issue close <n> --reason "not planned").SKIP- row to the tracker when the reason is worth keeping (section 13, Recorded skips)..cursor/rules/tracker-issues.mdc says: a new ID in the right section, never reused; Decision Accepted (<date>): the owner accepted it from #<n>; Status Not started; Source linking the issue. Start the issue's title with the ID (gh issue edit <n> --title "TON-31: …"), so the next sync adopts it and keeps the reporter's text above its block. Then bring the README roadmap and the Lightroom comparison into step (scripts/roadmap-sync.py, .cursor/rules/roadmap-and-comparison.mdc), commit, push, and sync the issues (scripts/tracker-issues.py, then --apply) only when .cursor/rules/main-branch.mdc allows it. Reply to the reporter: it's on the roadmap as the row, and the app's Your Reports shows it as Tracked as that ID.follows:<ID> (create the label the first time, gh label create follows:<ID> --color ededed), and reply that it's tracked as the row (link its issue) and that the reply will come there when it lands. It stays open for the person who asked.Answer it gives a question to a chat of its own. Answer from what the README, the docs, the manual and the code say, in the owner's voice, and close the issue as completed once it's answered. If the answer is "not yet" and it's worth doing, record what you found with --waiting you, so the owner can make it a suggestion; if it describes a bug, say so and let the owner triage it again.
Its brief is bug-agent.md: claim, read, reproduce when it isn't obvious, fix in a worktree of its own, push to main through the push gate, reply with the cause, the fix and the release, and close. It records every step with room.py note and syncs the room, so the Bugs tab shows each bug's track: Reported, Triaged, Reproduced (or not needed), Fixed, On main, Replied, Released. It stops and asks, flagged on the board, when the fix would change how existing edits render, when another session is rewriting the same code, while the release room is approved or releasing, and when it needs something only the owner has.
The expected release is the upcoming version from scripts/release-status.py, or the one after it when the release room has already checked a candidate that doesn't hold the fix. A fix that changes nothing in the app (the site, the cask, the docs) is out once it's on main.
When a release has shipped (room.py status shows fixes Released in it), reply on each report it fixed, in a sentence: "Redlamp 0.2.6 is out with this fix: Check for Updates… in the Redlamp menu installs it." Record it with room.py note <n> --out-url <the comment's URL>. A fix the room expected in that release but which isn't in it shows as missed: reply with a correction naming the release it comes in now, and record the new one in --release. The release room's last step brings this room up to date.
room.py watch syncs every 5 minutes (--every <seconds>); it only reads GitHub and git, so the owner can leave it running in a Terminal tab, or an agent in a background shell. A sync rewrites the room only when something changed, or when its last check is over four minutes old. A sync takes about a minute on this Mac, most of it git.
Replies on GitHub go out under the owner's account. Take his voice from his own messages (~/.cursor/projects/Users-pedrogomes-src-darkroom/agent-transcripts/*/*.jsonl, the user lines) and his comments on the issues: plain, first person, short, and kind to the person who took the time to report. Say what was wrong and what changed in words a photographer follows, not the code's. No superlatives, exclamation marks or em dashes, and none of the usual tells ("delve", "seamless"). British spelling (docs/brand/README.md). Claim only what the fix and its tests show.
When the owner wants it, a keeper chat can start each bug's agent without the click: room.py watch notices the new report, and the keeper hands a background agent the same brief.
The owner shapes the room as needs come up. A new list, state or button goes into the template, room.py and this skill in the same change, and the room picks it up. Commit them on main: Reports room: ….
© pdcgomes, MPL-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
SKILL.md and 3 other files in .cursor/skills/redlamp-reports of pdcgomes/redlamp.
Open the folder on GitHubat commit bc52683
Redlamp Reports next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| Redlamp Reports this skillpdcgomes/redlamp | 222 | — | ~2.7k | Automated safety check: Pass | MPL-2.0 | |
| Weavebench Cua ReproduceAMAP-ML/LongHorizon-Harness | 1.7k | — | ~1.6k | Automated safety check: Pass | MIT | |
| Evidence-Driven Testingmichaelshimeles/skills | 1.3k | 1 repos | ~3.9k | Automated safety check: Pass | None | |
| Create GitHub IssueNVIDIA/OpenShell | 16k | — | ~1.7k | Automated safety check: Pass | Apache-2.0 | |
| Triage IssuesClickHouse/clickhouse-java | 1.6k | — | ~904 | Automated safety check: Pass | Apache-2.0 | |
| Gentle AI Issue CreationGentleman-Programming/gentle-shell | 1.3k | — | ~2.5k | Automated safety check: Pass | Apache-2.0 |
AMAP-ML/LongHorizon-Harness
Reproduce CUA-Harness experiments on WeaveBench from a GitHub checkout.
michaelshimeles/skills
Records an annotated screen recording of the agent testing an app hands-on, then posts the video and a results summary to the PR and tracker issue.
NVIDIA/OpenShell
Create GitHub issues using the gh CLI. An agent skill from NVIDIA/OpenShell.
ClickHouse/clickhouse-java
Analyzes a single GitHub issue at a time. An agent skill from ClickHouse/clickhouse-java.
Gentleman-Programming/gentle-shell
Create and triage GitHub issues from repository evidence. An agent skill from Gentleman-Programming/gentle-shell.
termio-sh/termio
Diagnose a termio hang, beachball, crash, or 'it froze again' from the evidence macOS and termio actually leave behind — live process samples, crash and CPU-burn reports, the unified log, the…
pdcgomes/redlamp
Redlamp Bench, the way an agent asks the owner to do a step in another app and gets the results back without anyone moving files.
pdcgomes/redlamp
The Redlamp User Manual (docs/manual), a dense, book-style PDF for photographers, typeset from Markdown and a print style sheet by Chrome, with its ranges, defaults and shortcuts generated from the…
pdcgomes/redlamp
The promo studio, which makes Redlamp's short promotional videos for social feeds the way a small creative agency does - a brief, hooks and copy, a script on a music grid, visuals from the promo…
pdcgomes/redlamp
Map of redlamp.app, the Next.js website in web/: its stack and commands, its pages, the home page's sections with their anchors, components and content sources, the brand's Tailwind tokens and…
pdcgomes/redlamp
The blog room, where every redlamp.app blog post is tracked from idea to announcement in a canvas the owner marks as they post.
pdcgomes/redlamp
The press room, where every outlet Redlamp is pitched to (newsletters, Mac and photography sites, developer and open-source communities, lists and directories) is tracked from finding it to its…
Works with
Categories
The reports room, where every report people file on GitHub (Report a Bug and Send Feedback from the app, and issues opened by hand) is followed from arrival to the release that ships its fix, in a…. Redlamp Reports is an agent skill from pdcgomes/redlamp. The reports room, where every report people file on GitHub (Report a Bug and Send Feedback from the app, and issues opened by hand) is followed from arrival to the release that ships its fix, in a canvas the owner triages from.
Redlamp Reports fits situations like: feedback comes in; closing a report; applying the owners triage decisions; replying to a reporter.
Run `npx skills add pdcgomes/redlamp --skill redlamp-reports -a claude-code`. Or copy the skill folder (.cursor/skills/redlamp-reports in pdcgomes/redlamp) into .claude/skills/redlamp-reports in your project. Claude Code loads it when a task matches its description.
Run `npx skills add pdcgomes/redlamp --skill redlamp-reports -a codex`. Or copy the skill folder (.cursor/skills/redlamp-reports in pdcgomes/redlamp) into .agents/skills/redlamp-reports in your project. Codex loads it when a task matches its description.
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add pdcgomes/redlamp --skill redlamp-reports -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/redlamp-reports, .gemini/skills/redlamp-reports, .github/skills/redlamp-reports and .opencode/skills/redlamp-reports in your project.
Going by SKILL.md and its folder, Redlamp Reports needs Python for the scripts in its folder and the command-line tools its instructions call (gh and python3). Our summary lists: Python 3.
SKILL.md contains no URLs. Its commands use gh, which can reach the network depending on how they are called. This is read from the text; nothing was executed.
Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.
Redlamp Reports is published under the MPL-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
About 2.7k tokens (SKILL.md is roughly 11k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.
Skills that share tags, products or a category with Redlamp Reports: Weavebench Cua Reproduce (AMAP-ML/LongHorizon-Harness, 1.7k stars), Evidence-Driven Testing (michaelshimeles/skills, 1.3k stars), Create GitHub Issue (NVIDIA/OpenShell, 16k stars) and Triage Issues (ClickHouse/clickhouse-java, 1.6k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
pdcgomes (a GitHub user) maintains it in pdcgomes/redlamp, which has 222 GitHub stars. The repository holds 11 skills in this directory. The repository was last updated on October 10, 2026.
Source: pdcgomes/redlamp on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.