Toolkit
The machinery around the paper: LaTeX build/debug/lint, equation-to-code sync, publication figures, rebuttal triage and response, experiment tracking, cluster submission, artifact packaging.
/plugin install toolkit@research-agora
Checks against ground truth
Runs a tool or script and compares against something outside itself.
| Skill |
Description |
/code-simplify |
Analyze and refactor Python codebases to remove dead code, eliminate duplication, and simplify complexity |
/latex |
Build, debug and lint a LaTeX paper |
/latex-sync |
Keep a paper's equations and the code implementing them in agreement, via the latex-code-sync CLI |
/rebuttal |
Decode what reviewers actually want, then write the response |
Checks against a rubric
Applies a stated standard. You are the oracle.
| Skill |
Description |
/agora-feedback |
Opt-in, review-gated usage feedback for Research Agora skills (RFC-0001) |
/artifact-packager |
Use this agent to prepare ML code/data/models for public release with comprehensive checklists. Activates when asked to "package artifacts", "prepare release", "reproducibility checklist", "code release", or "prepare camera ready" |
Produces something for you to check
Generates an artifact or a candidate. Verifying it is your job.
| Skill |
Description |
/experiment-tracker |
Sync ML experiment results to paper drafts |
/figures |
Make publication figures for ML papers, in TikZ or matplotlib |
/htcondor |
Generate HTCondor submission files and wrapper scripts for ML research jobs |
License
MIT
1---2name: toolkit3description: Toolkit4---5# Toolkit67The machinery around the paper: LaTeX build/debug/lint, equation-to-code sync, publication figures, rebuttal triage and response, experiment tracking, cluster submission, artifact packaging.89```bash10/plugin install toolkit@research-agora11```1213## Checks against ground truth1415Runs a tool or script and compares against something outside itself.1617| Skill | Description |18|-------|-------------|19| `/code-simplify` | Analyze and refactor Python codebases to remove dead code, eliminate duplication, and simplify complexity |20| `/latex` | Build, debug and lint a LaTeX paper |21| `/latex-sync` | Keep a paper's equations and the code implementing them in agreement, via the latex-code-sync CLI |22| `/rebuttal` | Decode what reviewers actually want, then write the response |2324## Checks against a rubric2526Applies a stated standard. You are the oracle.2728| Skill | Description |29|-------|-------------|30| `/agora-feedback` | Opt-in, review-gated usage feedback for Research Agora skills (RFC-0001) |31| `/artifact-packager` | Use this agent to prepare ML code/data/models for public release with comprehensive checklists. Activates when asked to "package artifacts", "prepare release", "reproducibility checklist", "code release", or "prepare camera ready" |3233## Produces something for you to check3435Generates an artifact or a candidate. Verifying it is your job.3637| Skill | Description |38|-------|-------------|39| `/experiment-tracker` | Sync ML experiment results to paper drafts |40| `/figures` | Make publication figures for ML papers, in TikZ or matplotlib |41| `/htcondor` | Generate HTCondor submission files and wrapper scripts for ML research jobs |4243## License4445MIT