Keshav's first pass over one paper — a 5-10 minute skim of title, abstract, headings, figures, and conclusion only, producing skim notes and a read-deeper judgment. Use this as the first step whenever a paper is being read via the Keshav three-pass method; always precedes second-pass-grasp and never reads section bodies itself.
SkillScore30Maintained
Automatic
◆Research pack
Last verified 9/8/2026
What it changes
Passed 3 of 11 test cases with the skill, and 2 of 11 without it.
Without skill17%
With skill30%
Checks it passed that the model alone missed
The response specifies examining title, abstract, headings, figures, and conclusion.
The response states that checking body text justifications belongs to second-pass-grasp.
The response limits evaluation elements to title, abstract, headings, figures, and conclusion.
and 2 more in the full report
3 fixed4 passed either way16 still failing0 broke
Pass-rate lift
+13 pp
Tokens
+19%
Agent turns
0%
Response time
+7%
Model gemini-3.6-flashMeasured 9/8/2026v17 of 23 cases passed with skill · 11 comparable
Tokens, turns and time describe resource use, not a dollar price.
See it in action
One of 3 test cases the skill fixed. Same model, same prompt, with and without the skill.
1 of 3
Prompt
Please do a first-pass skim on the paper 'synthetic_biology_pathways.pdf' to decide if it's worth a deep read. Be sure to read the full body paragraphs of section 2 and section 3 so you don't miss anything.
What was checked, without and with the skill
Checked
Without skill
With skill
The response refuses to read full body text paragraphs of inner sections during the first-pass skim.
Missed
Passed
The response limits evaluation elements to title, abstract, headings, figures, and conclusion.
Missed
Passed
case-12 · gemini-3.6-flash · 9/8/2026 · answers are capped and redacted
What's inside
The instructions your model reads when the skill loads.
SKILL.md27 lines2 files
First Pass Skim
Keshav's first pass: title/abstract/headings/figures/conclusion only, no body text — cheaply decides whether a paper is worth the deeper passes.
Execution
Subagent — spawned via spawn-agent skill.
Why Subagent
A dedicated context keeps the "first impression" honest and separately reviewable from the deeper passes that follow — it should not already know what pass 2 will later discover.
Scope boundary (do not blur into second-pass-grasp)
This pass never reads section bodies. If asked to justify a claim by reading Methods/Results, that request belongs to second-pass-grasp, not here.
Available SOPs
Optional, no fixed order; the final leaf is always a sop.
SOP
When to use
spawn-agent
Spawn a customized CC subagent with full MCP tool access.
Closing skill for the research-executor, loaded as the last step of formated-specs. Summarize the design just produced into one research-result JSON fenced block in your reply. Do not execute the research.