Strategic critique of the overall research state. Reads researchlog.md (with attention to Established Results, Working Claims, Dead Ends), plan.md, and may spot-check artefacts. Identifies strategic drift, plan-viability issues, coherence problems in Established Results, under-sourced claims, and citation gaps.
Perform an analytical derivation for a target claim or open question. Works only from what the main agent dispatches. Writes derivations/D-NNN.md with the derivation, stated assumptions, and intra-derivation sanity checks.
Synthesise answer.md — the final research deliverable — from Established Results and supporting artefacts. Reads researchlog.md, problem.md, plan.md, and selectively reads derivations/, computations/, critiques/, references/ as needed. Writes plain-markdown answer.
Draft or update plan.md — a numbered research strategy for the problem. Reads problem.md, survey.md, and researchlog.md. On re-run, marks completed steps, drops obsolete steps, and revises remaining steps based on progress.
Research-landscape survey for a problem in theoretical physics or mathematics. Reads problem.md and runs web/arxiv search; produces survey.md with background, known approaches, pitfalls, and key references. May fetch papers into references/.
Perform symbolic and numerical computation for a target claim. Default backend SymPy + NumPy. Runs symbolic and numerical paths together when both are feasible and flags disagreements. Writes computations/C-NNN.md (narrative) and computations/C-NNN.py (executed script).
Strategic critique of the overall research state. Reads researchlog.md (with attention to Established Results, Working Claims, Dead Ends), plan.md, and may spot-check artefacts. Identifies strategic drift, plan-viability issues, coherence problems in Established Results, under-sourced claims, and citation gaps.
Perform an analytical derivation for a target claim or open question. Works only from what the main agent dispatches. Writes derivations/D-NNN.md with the derivation, stated assumptions, and intra-derivation sanity checks.
Synthesise answer.md — the final research deliverable — from Established Results and supporting artefacts. Reads researchlog.md, problem.md, plan.md, and selectively reads derivations/, computations/, critiques/, references/ as needed. Writes plain-markdown answer.
Draft or update plan.md — a numbered research strategy for the problem. Reads problem.md, survey.md, and researchlog.md. On re-run, marks completed steps, drops obsolete steps, and revises remaining steps based on progress.
Research-landscape survey for a problem in theoretical physics or mathematics. Reads problem.md and runs web/arxiv search; produces survey.md with background, known approaches, pitfalls, and key references. May fetch papers into references/.