Profile My skills Saved Collections Edit profile Submit a skill

← all publishers

EmmanuelRTM

@emmanuelrtm source repo

11 published skills

  1. ▌
    Discover · emmanuelrtm
    Run (or re-run) Stage 2 only — Discovery: Ideator plus Parallel Explore-Exploit branches (Solve → Evaluate → Audit → Rank+Select → distilled feedback), for the configured number of iterations. Requires brief.md from /investigate.
    0
    installs
  2. ▌
    Research · emmanuelrtm
    Run the full autonomous research pipeline end-to-end: investigate the literature, discover a solution via parallel explore-exploit branches, write an evidence-tagged paper, and verify every claim. The ScientistOne pipeline as one command.
    0
    installs
  3. ▌
    Run Status · emmanuelrtm
    Show the state of the active research run: stages completed, branch scoreboard, selected solution, grounding ratio, artifacts present, and the suggested next stage command.
    0
    installs
  4. ▌
    Investigate · emmanuelrtm
    Run (or re-run) Stage 1 only — the Problem Investigator: citation graph, elite pool, deep reading, and the grounded Research Brief. Operates on the active run, or scaffolds a new one if a topic is given.
    0
    installs
  5. ▌
    Write Paper · emmanuelrtm
    Run (or re-run) Stage 3 only — the Paper Writer pipeline: CONCEIVE (LLM) → GROUND (deterministic script) → CRITIC (fresh-context LLM) → RESOLVE+COMPOSE (LLM) producing the evidence-tagged paper draft. Requires best/SELECTED.json from /discover.
    0
    installs
  6. ▌
    Verify Claims · emmanuelrtm
    Run (or re-run) Stage 4 only — the Claim Verifier: extract every claim from the draft, verify each against its declared evidence source (numerical ±5%, citation entailment, method-code alignment), refine failures, finalize the paper, and run the Chain-of-Evidence audit. Requires paper/draft.md.
    0
    installs
  7. ▌
    Paper To Latex · emmanuelrtm bundle
    Convert the finalized paper (final/paper.md + bibliography.jsonl) into LaTeX using the bundled template, and compile to PDF when a LaTeX toolchain is available. Optional post-processing after /verify-claims.
    0
    installs
  8. ▌
    Distill Feedback · emmanuelrtm
    Template for distilling one PEE iteration's results into feedback for the next Ideator round (the Rank+Select → Ideator feedback edge). Used by the main agent after ranking an iteration's branches.
    0
    installs
  9. ▌
    Evidence Tagging · emmanuelrtm
    Evidence-tag grammar and tagging discipline for paper artifacts. Preloaded into every agent that writes, audits, verifies or repairs a paper artifact (paper-writer, paper-critic, claim-verifier, refiner); defines how every factual claim is bound to a workspace artifact (Chain-of-Evidence).
    0
    installs
  10. ▌
    Citation Provenance · emmanuelrtm
    Literature retrieval and provenance protocol for the problem-investigator. Defines the citation-graph rounds, elite-pool filtering, the degradation ladder for restricted networks, and bibliography.jsonl discipline.
    0
    installs
  11. ▌
    Evaluation Protocol · emmanuelrtm
    Official scoring and audit protocol for Parallel Explore-Exploit branches. Preloaded into the ideator, solver, evaluator, auditor, and ablation-analyst agents; defines eval.json, the audit checklist, and what counts as a specification violation.
    0
    installs