Mmcp.market

coauthor-brief skill

by pedrohcgs·pedrohcgs/claude-code-my-workflow·1.6k stars·MIT

Generate a co-author / collaborator handoff brief for a multi-author, multi-machine project — summarizing what changed since the last brief (git delta), the current state of each artifact (manuscript, analysis, slides), open questions, how to reproduce locally, and any restricted-data access steps. Use when user says "coauthor brief", "handoff brief", "bring my coauthor up to speed", "what changed since last week", "onboard a collaborator", "write a handoff for [name]", or before sending a co-author the repo. NOT a commit or a checkpoint — it is the cross-machine, cross-person summary `meta-governance.md` only partially covers.

A100/100content scan

Is the coauthor-brief skill safe?

Clean: nothing in its files matched our rules. We read 1 file in the folder on 2026-09-28.

No findings.

Install the coauthor-brief skill

A skill is a folder. Copy it into your agent's skills folder and the agent loads it when the task matches its description.

git clone --depth 1 https://github.com/pedrohcgs/claude-code-my-workflow.git /tmp/claude-code-my-workflow
mkdir -p ~/.claude/skills
cp -r /tmp/claude-code-my-workflow/.claude/skills/coauthor-brief ~/.claude/skills/coauthor-brief
available in every project

In the Claude apps, zip the folder and upload it from the Skills settings. The folder on GitHub

The instructions your agent would load

SKILL.md as published, without the frontmatter. Read it on GitHub

/coauthor-brief — Collaborator Handoff Brief

Produce a single Markdown brief a co-author (or your future self on another machine) can read in a few minutes to know what changed, where each artifact stands, what's blocked, and how to run the pipeline locally — including restricted-data access steps a new collaborator needs. meta-governance.md covers the memory side of cross-machine work (what syncs via git, what stays in gitignored native auto memory); this skill covers the human side: the per-person, per-session handoff.

Core principle: /checkpoint is for you resuming; /coauthor-brief is for someone else starting. The first answers "where am I?"; the second answers "what do I need to know to take over a piece of this?"

When to use

  • Before sending a co-author the repo (or a PR / branch) and you want them oriented, not archaeologizing the git log.
  • Onboarding a new RA / collaborator who needs the reproduce-locally and restricted-data steps in one place.
  • Periodic sync on a long multi-author project — "here's the delta since the last brief."
  • Cross-machine handoff — finishing on the office desktop, picking up on the laptop, or handing the analysis half to a co-author who runs Stata while you run R.

When NOT to use

  • For your own resume-point — use /checkpoint.
  • For distilling a noisy session before auto-compact — use /compress-session.
  • For the commit message itself — that goes through /commit.

Phases

Phase 0 — Determine the "since" point

Resolve the delta window, in this priority order:

  1. --since if given — a git tag, an ISO date, or Ndays (e.g. --since v1.2, --since 2026-05-01, --since 14days).
  2. Else, the last brief — ls -t qualityreports/handoffs/.md | head -1; use the date in its filename as the floor.
  3. Else, fall back to 14 days and say so in the brief (an unbounded git log is noise, not signal).

Echo the resolved window back before gathering ("Brief covers changes since …").

Phase 1 — Gather the delta + project state

Read-only collection. Skip any source that doesn't apply (R-only, Stata-only, no slides) rather than fabricating.

  1. Git delta — git log --oneline --since= (or ..HEAD), git diff --stat ..HEAD, git branch --show-current, and git log --all --oneline -15 to see co-authors' parallel branches. Group commits by area (manuscript / analysis / slides / infra).
  2. Reproducibility status — if a passport exists at quality_reports/passports/.yaml, read the PASS / FAIL / EXPLAINED / STALE / UNVERIFIED roll-up (audit-reproducibility, replication-protocol.md). Note whether the analysis is replication-ready or has open FAILs.
  3. Open plan items — most recent qualityreports/plans/.md (status + any "Open questions" / "Next" lines) and the latest qualityreports/sessionlogs/*.md blockers.
  4. Environment / lockfiles — locate the env capture a collaborator needs: renv.lock (R), requirements.txt / environment.yml / uv.lock (Python), a Stata version line + .do ssc install list, or whatever /capture-environment wrote. Record the exact restore command.
  5. Restricted-data steps (skip if --no-data-section) — if the repo touches confidential/restricted data, read confidential-data.md and summarize the access path a new collaborator needs (DUA / IRB approval, secure-enclave or openICPSR-restricted credentials, where the data lives, what is git-ignored vs committed). Never copy actual restricted values, paths to live extracts, or credentials into the brief — describe the process to obtain access, not the data.

Phase 2 — Write the brief

Use the template below. Keep it tight (~1–2 screens). Concrete path:line pointers beat prose.

Label every status statement measured or inferred. Measured means a command was run and what you are reporting is its output — name the command. Inferred means you believe it from reading the code, the diff, or a run from some earlier day. Both belong in a brief; mixing them silently does not, because the reader cannot tell which claims they can build on without re-checking. Write replication-ready (measured: Rscript scripts/R/00runall.R — 0 FAIL), not replication-ready.

Disposition delegated work item by item. Anything handed to a co-author, an RA, or an agent gets its own line and its own state — done (with the evidence) / in progress / blocked (on what) / not started. "Mostly done" is not a disposition: it hands the reader the job of finding the missing piece, which is precisely the job the brief exists to do.

---
date: YYYY-MM-DD
for: [collaborator name or "all coauthors"]
since: [tag | date | Ndays]
branch: [current branch]
---

# Co-Author Brief — [project / paper short name]

## What changed since [since-point]
[3–8 bullets, grouped by area. Each: what changed + why it matters to a reader, not raw commit subjects.]
- **Analysis:** re-ran the main specification on the de-duplicated sample (533 fanned-out rows dropped); headline estimate now −1.19 — see `scripts/R/03_analyze.R:147`.
- **Manuscript:** Table 2 + §4.2 rewritten to match; Figure 3 regenerated.
- **Slides:** untouched.

## Current state of each artifact
| Artifact | Path | State | Evidence (measured / inferred) |
|---|---|---|---|
| Manuscript | `manuscript.tex` | drafting §5 | inferred — §5 read, not compiled since the Table 2 rewrite |
| Analysis | `scripts/R/` | replication-ready | measured: `Rscript scripts/R/00_run_all.R`, passport 11 PASS / 1 EXPLAINED / 0 FAIL |
| Slides | `Slides/` | current | measured: `git log --since=<point> -- Slides/` returns nothing |

## Delegated work — one line per item
[Everything handed to a co-author, an RA, or an agent, dispositioned individually. No roll-ups.]
| Item | Owner | State | 

Phase 3 — Save and summarize

  1. Write to qualityreports/handoffs/YYYY-MM-DDcoauthor-brief.md (create qualityreports/handoffs/ if missing). If --for is set, suffix the slug (…coauthor-brief_.md).
  2. Print to chat: saved path, the resolved since-window, counts (commits summarized / open questions / artifacts), and whether the data section was included or skipped.

Output / report format

A single Markdown handoff doc at qualityreports/handoffs/YYYY-MM-DDcoauthor-brief.md matching the Phase 2 template, plus a one-line chat summary. No edits to any other file.

Exit behavior

  • Brief written: exit 0 with the saved path and summary line.
  • No "since" resolvable and no git history in window: still write the brief with an explicit "no changes in window" note rather than failing — a co-author starting fresh still needs the state + reproduce + data sections.
  • Restricted-data repo detected but confidential-data.md missing: write the brief, leave the data section with a TODO: complete per your DUA placeholder, and warn — do not guess the access process.

Flags

  • --since — Baseline to diff against — a git tag, an ISO date, or Ndays (e.g. 14days). Default: the previous brief in quality_reports/handoffs/, else 14 days (Phase 0).
  • --for — Tailor the brief to a specific collaborator (e.g. surface the restricted-data access steps they still need).

Cross-references

  • .claude/rules/meta-governance.md — the cross-machine memory model (what syncs, what stays local) this brief operationalizes for people.
  • .claude/skills/capture-environment/SKILL.md — produces the lockfiles the "Reproduce locally" section points at.
  • .claude/skills/checkpoint/SKILL.md — the self-resume companion (this skill is the other-person handoff).
  • .claude/skills/compress-session/SKILL.md — distil a noisy session before compaction (orthogonal; run before briefing if context is fat).
  • .claude/rules/confidential-data.md — restricted-data handling; the data section summarizes its access steps without exposing data.

What this skill does NOT do

  • Push, PR, or merge. It writes a doc. Branch, commit and PR are /commit's job; a merge is the user's call.
  • Run the pipeline or audit numbers. It reports passport status if one exists; it does not re-run analysis or re-verify claims — that's /audit-reproducibility.
  • Capture the environment. It locates and links existing lockfiles; generating them is /capture-environment.
  • Expose restricted data. It describes the access process only — never copies confidential values, live data paths, or credentials into a brief that may be emailed or committed.

More skills from pedrohcgs/claude-code-my-workflow

  • Aadjudicate-reviewTurn an incoming set of findings — from an AI reviewer, a referee report, a code review, a linter, or a second model — into verified fixes, without letting a confident misread damage correct work. Every finding is a CANDIDATE until checked against the actual source. Use whenever you receive review comments, audit findings, or a critique you did not write yourself, especially when the reviewer is a model or when the volume is too large to check by feel.
  • Aaudit-reproducibilityEnforce the replication-protocol.md rule by cross-checking numeric claims in a manuscript against the actual R / Stata / Python outputs. Report PASS/FAIL per claim against tolerance thresholds. Use before submission and before releasing a replication package.
  • Ablast-radiusBefore and after changing anything shared — a function's return value, a signature, a schema, a label set, a config default, a constant, a file format — find every consumer and actually run them. Catches the change that looks purely additive but silently breaks a contract in a file you never opened. Use when editing shared code, adding a field/column/return element, renaming, changing units or defaults, or touching a pipeline that produces reported numbers.
  • Acapture-environmentSnapshot the computational environment for a replication package — detects the analysis stack (R / Stata / Python) and emits the right lockfiles (renv.lock + sessionInfo.txt, requirements.txt / environment.yml / uv.lock, Stata version + ado package list), records seeds and RNG kind, optionally writes a pinning Dockerfile, and produces a paste-ready "Computational requirements" block. Use when user says "capture the environment", "snapshot my dependencies", "pin the versions", "make a renv.lock / requirements.txt", "make this byte-reproducible", or before releasing a replication package to openICPSR / the AEA Data Editor.
  • AchallengeStress-test a finding against the choices you did not make. Enumerates the discrete forks a competent analyst could have taken (measure definition, sample filter, control set, clustering level, weighting, functional form), runs the specification grid, and reports the distribution rather than a point estimate — then attacks the identifying assumption with named, computable sensitivity statistics. Use when the user says "is this robust", "challenge this result", "specification curve", "multiverse", "how sensitive is this", "what if I'd used a different measure", "stress-test my estimate", or before a result becomes a headline claim. NOT a reviewer of prose or code — it challenges the CLAIM.
  • AcheckpointSave a structured state snapshot before stopping or handing off. Captures the active plan, recent decisions, file pointers (with line numbers), open questions, and the next 1–3 actions into a checkpoint file under `quality_reports/checkpoints/`. Optionally proposes `[LEARN]` entries to add to MEMORY.md. Use when user says "checkpoint", "save state", "snapshot before I stop", "where am I", "wrap up the session for handoff", or before a long break / model switch / collaborator handoff. Companion to (NOT replacement for) the narrative session-log workflow.
  • AcommitCommit the current work — runs the quality, consistency and passport gates, branches off main if needed, stages specific files, and writes a commit whose subject states what is now true. Pushes and opens a pull request only with --pr or when the user asks; never merges — a merge happens only when the user explicitly says to merge. Use ONLY on explicit commit intent — user says "commit", "let's commit this", "open a PR", or prefixes with `/commit`. Do NOT auto-invoke on vague end-of-task phrases ("we're done", "wrap up") — those require explicit confirmation first. Never force-pushes or skips hooks.
  • Acompile-latexCompile a Beamer LaTeX slide deck with XeLaTeX (3 passes + bibtex). Use when user says "compile", "build the slides", "rebuild the PDF", "run latex", "render the tex", or asks why a `.tex` file isn't producing a PDF. Operates on `Slides/*.tex`.
  • Acompress-sessionDistill the current conversation into a structured note (decisions made, open questions, file pointers with line numbers, next 1–3 actions) and save to `quality_reports/session_logs/` before auto-compression. Differs from `/checkpoint` (explicit stop-point snapshot) and from auto-compaction (which truncates rather than distills). Use when context is approaching auto-compact threshold, when a long pipeline has accumulated many decisions, or when the user says "compress", "distil this session", "before we hit auto-compact", "structured handoff before context resets".
  • Acontext-statusShow current context status and session health. Use to check how much context has been used, whether auto-compact is approaching, and what state will be preserved.
  • Acreate-lectureCreate a new Beamer lecture `.tex` from source papers and materials, with notation consistency checks and the project's preamble wired in. Use when user says "create a lecture on X", "new lecture from these papers", "start a deck on topic Y", "scaffold a new Beamer file", "build me a lecture from these PDFs". Scaffolds the full deck — NOT for compiling existing `.tex` (use `/compile-latex`).
  • Acredible-claimsResearch-brief + claim-record discipline for delegated or AI-assisted research work. Use when starting any substantive research task or long autonomous run (write the brief first), and when reporting results that will support a claim in a paper or decision (produce the claim record). Keeps faster execution from being confused with credible evidence.

All agent skills → · MCP servers