Figure Composer

What it does

When your agent uses it

• Tasks that involve Subagents

Example prompts

• “/figure-composer”

Requirements

• Python 3

Workflow steps

5 steps, taken from the step headings in SKILL.md.

1. 1Where this sits
2. 2Narrative → panel outline
3. 3Fan-out (one sub-agent per panel)
4. 4Compose
5. 5Adversarial self-review loop (two-tier, design rules held fixed)

What it can do on your machine

Read from SKILL.md and the folder at commit fb309c3. It shows what the files ask for, not the result of running them.

• Tool permissions

  Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

  From allowed-tools in the SKILL.md frontmatter.

• Runs code

  Ships script files (Python), which the agent can run.

  From the folder's file list and the shell code blocks in SKILL.md.

• Network

  No URLs in SKILL.md.

  From URLs in SKILL.md, links to its own repository left out.

• Credentials

  Names no API keys, tokens, secrets or passwords.

  From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Figure Composer loads about 1.7k tokens when it runs. Until then it costs about 169 tokens; SKILL.md has 602 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it

~169

When it runs · the whole SKILL.md, loaded when a task matches

~1.7k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from JimLiu/science-skills at commit fb309c3, republished under its Apache-2.0 licence (© JimLiu). 602 words, ~1,673 tokens.

name

figure-composer

description

Compose one publication-grade multi-panel figure. Entry from a one-line claim + data refs, OR from an existing figure via `derive_outline(png)`. Runs a per-figure loop: outline (12-col grid, per-panel ask + label_budget) → fan-out one sub-agent per panel (each loads `figure-style`) → tile + stamp letters → adversarial composite review with two-tier feedback (Tier-1 outline_revisions / Tier-2 per-panel violations) → regen affected panels, ≤3 rounds. Loads panel_task / compose_figure / compose_crops / composite_review_task / derive_outline into the kernel. For one standalone plot use `figure-style`; for whole-paper figure ordering use `paper-narrative`.

license

Apache-2.0

{"claim":"…", "width_mm":180, "ncol":12, "row_heights_mm":[40,60,46,52],
 "panels":[
  {"letter":"a","role":"schematic","row":0,"col":0,"colspan":12, "chart_family":"schematic overview", "message":"…", "data_vid":null, "ask":"…"},
  {"letter":"b","role":"primary",  "row":1,"col":0,"colspan":7,  "chart_family":"scatter + trend", "message":"…", "data_vid":"…", "ask":"…"},
  …]}

requests = [{"name": f"panel-{L}", "task": tasks[L],
             "output_schema": {"type":"object","properties":{"figure_filename":{"type":"string"}},
                               "required":["figure_filename"]}}
            for L in letters]   # no "profile" key — default agent profile
descs = host.delegate(requests, wait=False)

out_path, (W, H) = compose_figure(outline, panel_paths, "fig.png")
for L, box in compose_crops(outline).items():
    host.view_image("fig.png", crop=box)

loop (max 3 rounds, floor 5→4→3):
  review = delegate(composite_review_task(composite_vid, outline, rules_vid, prev_vid, round, floor))
  if review.editor_verdict in {accept, minor_revision} and 0 BLOCKER and ≤2 MAJOR: break

  # TIER 1 — outline-level
  if review.outline_revisions:
      apply revisions to `outline` (geometry, row-header titles, label_budget, panel set)
      affected = apply_outline_revisions(outline, review.outline_revisions)
  else:
      affected = set()

  # TIER 2 — panel-level
  fixb = group_fixes_by_panel(review)       # BLOCKER/MAJOR only
  regen = affected | set(fixb)              # only these panels regenerate
  re-delegate each L in regen with panel_task(outline, L) + fixb.get(L,"") +
      "do not over-correct: where the previous version was correct, keep it"
  recompose

© JimLiu, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 1 other file in skills/figure-composer of JimLiu/science-skills.

• SKILL.md
• kernel.py

Open the folder on GitHubat commit fb309c3

Used in 2 other repositories

We found 2 copies of this SKILL.md (exact, near-identical or edited) in other folders, from 2 other GitHub owners. This page covers the copy in JimLiu/science-skills, which our catalogue first saw on October 7, 2026.

• BioTender-max/awesome-bio-agent-skillsskills/claude-science/figure-composer
• PKU-YuanGroup/OpenAI4Sskills/figure-composer