← ClaudeAtlas

reevaluate-strategylisted

Learn from accumulated outcomes — read the funnel, rejections, interview transcripts, and wins, then recommend concrete tuning to targeting signals, comp anchoring, fit calibration, channel mix, and writing-style. Triggered by outcome thresholds or an explicit review request.
CodesWhat/careerrat · ★ 3 · AI & Automation · score 72
Install: claude install-skill CodesWhat/careerrat
# reevaluate-strategy Use this skill when the user asks why they're getting filtered, to review their strategy, to re-rank sourced roles, or "what should I change" — or when `track-outcomes` trips a threshold (see the Reevaluation Contract in `AGENTS.md`). This is the self-tuning half of the loop: it turns accumulated outcomes into concrete, recommended adjustments. > **Runs under AGENTS.md.** These contracts bind without being restated here: Privacy Invariant (`current_base` never outbound), Honesty Firewall, Placeholder/Bracket Ban, Gate Write-back, Domain-Neutral Rule, Browser Automation Contract, Activity Pulse logging, Tracker verify+snapshot, and Sent-Clears-Draft. Inline reminders at point-of-use are intentional; standalone restatements point back to the relevant AGENTS.md section. > **Agent voice.** Read candidate modes through the shared DB-first accessor (`modes.agent_voice`, default `standard`) before producing strategy recommendations. Apply the register from AGENTS.md#mode-switches. `exec-summary` = top 2–3 recommended changes as bullets + threshold that tripped; `standard` = findings summary + recommendations as scannable bullets; `technical` = funnel analysis + per-signal breakdown + recommended changes; `verbose` = full analysis including win/rejection patterns, confidence levels, and proposed gate write-backs. --- ## STEP 0 — GATE CHECK (trigger confirmation) **Trigger sources.** This skill is entered either: - Explicitly by the user ("why am I getting