tjhavranek
UserThree Claude Code skills for working with Codex CLI: codex-bridge (one-shot Codex calls), mad-build (Claude+Codex collaboration with cross-review), and mad-research (three-stream adversarial audit of papers, grants, reports with anonymized cross-critique and fresh-Codex synthesis).
Categories
Indexed Skills (4)
mad-research
Three-role-stream adversarial audit of a research document — paper, grant, referee report, preprint. Quote+page-grounded criticism, anonymized cross-critique, fresh-context Codex synthesis against a locked rubric, minority report preserved. Two modes: default methodology audit, and an opt-in Bayesian Mode for evaluating the truth of a contested empirical claim. Can also run in an opt-in Claude-only configuration (all streams + synthesis via fresh-context Claude subagents; single-provider, not cross-model). Triggers: - "MAD-research <file>" - "Run MAD on <file>" (when the file is paper-like: .pdf, manuscript, grant, preprint) - "stress-test this paper" / "stress-test this grant" - "referee report on this" - "audit this manuscript" - "MAD-audit <file>" - "MAD-research in Bayesian Mode" / "evaluate the empirical claim X with Bayesian discipline" (opt-in Bayesian Mode) - "MAD-research <file> Claude-only" / "run mad-research in Claude-only mode" (opt-in single-provider mode — all streams + synthesis via fresh Clau
codex-bridge
Lets Claude call the Codex CLI for a single ad-hoc task — no protocol, no rounds, no synthesis. Use this when you want a one-shot second opinion or a quick external draft without the full MAD scaffolding. Triggers: - "have Codex <verb> <thing>" (e.g. "have Codex draft X", "have Codex review file Y") - "ask Codex to <verb>" - "use Codex for <task>" - "run Codex on <file>" - Explicit "codex-bridge" mention. For structured Claude + Codex collaboration (drafts → cross-review → revisions → merge), use the mad-build skill instead. For an adversarial audit of a research document, use the mad-research skill instead.
mad-build
Four-step Claude + Codex collaboration for producing artifacts (code, drafts, analyses). Independent drafts → cross-review → revisions → merge. Triggers: - "MAD-build <task>" - "have Codex help me build X" / "have Codex help me draft Y" - "competition agent on <task>" - "let Codex try this too" - "do this with both AIs" For an adversarial audit of a research document (papers, grants, referee reports), use the mad-research skill instead. For a one-shot Codex call without protocol, use the codex-bridge skill instead.
paper-workshop
CRUCIBLE — a Claude-only adversarial expert workshop that stress-tests a scientific paper and then, opt-in, rebuilds it. ACT I (TRIBUNAL): from the paper PDF (plus a few public cited works it fetches), a topic-adapted fleet of expert referee subagents — every contested claim argued from at least two COMPETING intellectual traditions, plus generalist seats for relevance and understandability — reviews the paper to the sentence, every comment grounded in an exact quote (no fabrication, no confidence scores) and independently re-checked from many angles before delivery. ACT II (ATELIER, opt-in): with the author's source + data + code, it implements the agreed changes as a tracked-changes redline plus a clean revised version, RE-RUNS the analysis under an execution-provenance wall (no number enters the paper unless a logged re-run produced it), and assembles a replication package — every change multi-angle-verified, nothing applied without the author's sign-off on anything that touches the record. Triggers: - "wo
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.