菜单
Print the current state of the autoresearch loop — what we've learned, what's being tested, what's pending review, what's next. Reads everything in .cold/ and outputs a one-screen dashboard. Use anytime to orient. No FoxReach calls — purely local read.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
Run per-lead web research on a candidates CSV, write one bundle.md per lead, and emit assignments.json splitting leads into A/B arms for experiments. Uses parallel sub-agents for batches of 10. Spawned by cold-experiment when the experiment design has personalization as the variable, or by the user saying "research these leads", "personalize this list", "build per-lead bundles".
Import a CSV of leads, score each against icp.md, drop misaligned ones, optionally enrich with per-lead web research, then push the survivors into FoxReach as a campaign. Replaces the v0 stub. Use when the user says "import leads", "score these prospects", "filter to my ICP", "build a campaign from this CSV".
Read the active experiment's results from FoxReach, run the statistical test, and propose a cold.md diff (or auto-commit if trust earned). Computes interested-reply rate per variant via foxreach inbox categorize-stats, applies a two-proportion z-test, checks guards (bounce rate, not-interested rate), and either writes .cold/proposed-diff.patch or applies it directly. Use 7+ days after /cold send when an experiment is active. Updates beliefs.md and decisions.md. Use when the user says "read the experiment", "/cold learn", "what did we learn", or "is the test done yet".
Push the active campaign to FoxReach — creates sequences, registers variants from cold-draft output, runs pre-flight checks (banned-phrase lint, audience check, plan-limit math, bounce-rate guard), and starts the campaign. Use after /cold leads + /cold draft. Replaces v0 stub.
Pull unread inbox replies from FoxReach, surface interested ones, draft responses per cold.md
Generate a digest of cold outreach performance - deliverability, replies, booked calls, domain health. Use when the user asks for a report, digest, summary, weekly recap, performance check, or asks "how did my campaigns do". Reads live metrics from the FoxReach API. Designed to be called by cron (daily or weekly) and post results to Slack or email. Part of the cold.md suite.
基于 SOC 职业分类
| name | cold-status |
| description | Print the current state of the autoresearch loop — what we've learned, what's being tested, what's pending review, what's next. Reads everything in .cold/ and outputs a one-screen dashboard. Use anytime to orient. No FoxReach calls — purely local read. |
The autoresearch dashboard. Use to orient yourself before kicking off the next step.
/cold <verb> command, to confirm prerequisites.cold/config.json.cold/beliefs.md.cold/decisions.md.cold/experiments/.active.cold/experiments/<id>/protocol.md.cold/experiments/<id>/decision.md (if concluded).cold/proposed-diff.patch (if pending review).cold/trust.json.cold/last-import.json, .cold/last-send.jsoncold.md, icp.md, offer.md (existence checks)If .cold/config.json doesn't exist:
cold.md not initialized. Run `/cold init` first.
Stop.
Read all the files listed in Inputs (silently skip missing ones).
═══════════════════════════════════════════════════════════════
cold.md autoresearch — status as of <YYYY-MM-DD HH:MM>
═══════════════════════════════════════════════════════════════
POLICY FILES
cold.md ✓ (last edit: <relative time>)
icp.md ✓
offer.md ⚠ missing — run /cold offer
CAMPAIGNS
Active campaign: cmp_abc123 ("Q2 SaaS founders")
Leads: 87 imported (142 scored, threshold 0.5)
Sent at: 2026-04-22 (7 days ago)
EXPERIMENTS
Active: 2026-W18-subject-statement-vs-question
Variable: subject pattern (Tier 1)
Arms: arm_a_statement vs arm_b_question (50/50)
Sample: 100/100 sent — ready to read ✓
Days: 7 of 7 elapsed
Last concluded: <none yet>
BELIEFS (current best-known patterns)
<none yet — first experiment in flight>
PENDING REVIEW
⚠ Proposed cold.md diff at .cold/proposed-diff.patch
From experiment: 2026-W18-subject-statement-vs-question
Outcome: arm_b_question wins (+3.0pp, p=0.04)
Review: cat .cold/proposed-diff.patch
Accept: git apply .cold/proposed-diff.patch && rm .cold/proposed-diff.patch
Reject: rm .cold/proposed-diff.patch
TRUST LADDER
Approved-diff streak: 0 / 3
Auto-commit: OFF (will unlock at streak 3)
PERSONALIZATION
Per-lead: enabled — 87 lookups cached at .cold/research/lead-personalization/
DECISIONS LOG (last 3)
- 2026-04-29 — Tier 1 / subject — pending review
- <older entries>
═══════════════════════════════════════════════════════════════
SUGGESTED NEXT
→ Review and accept the pending diff (above)
→ Then: /cold experiment to design Tier 2 (opener)
═══════════════════════════════════════════════════════════════
/cold experiment after /cold leads/cold leads + /cold experiment to start the first loop.cold/decisions.md for the breachPick the single most important next action based on state:
| Current state | Suggested next |
|---|---|
No cold.md, icp.md, offer.md | /cold init then /cold icp <url> |
| Has icp.md, no offer.md | /cold offer |
Has cold.md, no last-import.json | /cold leads --csv ./prospects.csv |
Has import, no last-send.json, no active experiment | /cold experiment |
| Has active experiment, no drafts | /cold draft |
Has drafts, no last-send.json | /cold send |
| Active experiment, days remaining | "Wait N more days, then /cold learn" |
| Active experiment, ready to read | /cold learn |
| Pending diff | "Review and accept/reject the proposed diff" |
| Diff resolved, no active experiment | /cold experiment (next tier) |
.cold/. (Reading is fine.)/cold report for FoxReach-side stats.✓ ⚠ ✗)./cold report