refactor: drop redundant Pilot experienceTracker param + tracker access cleanup

[DavertMik]

Follow-up to #84 (merged). Removes the remaining redundant tracker plumbing and tidies access.

Changes

• Pilot: same footgun as Navigator — its only caller passed the shared ExperienceTracker and Pilot already holds explorer, so the optional experienceTracker? param + || null was pure redundancy. Dropped the param and the field; getExperienceToc() derives it lazily from explorer.
• createAgentTools: took experienceTracker + a getState closure solely to build the learnExperience tool — redundant for Pilot/Tester/Captain/Navigator (all derived them identically from explorer), while Rerunner encoded "no experience tool" by omitting them. Replaced both with a single explicit withExperience flag (default on); the tracker + state are derived lazily inside the tool's execute (registering tools never touches getExperienceTracker, so minimal test explorers are safe). Rerunner passes withExperience: false. Removes the duplicated getState closure from all four call sites.
• Captain.getKnowledgeTracker() → explorBot.knowledgeTracker() (owned instance; works before the explorer starts).
• Type-only imports → import type (state-manager, captain, historian, navigator); dead ActionResult imports removed (explorbot, web-mode).

Verification

bun run lint clean; bunx tsc --noEmit at baseline (no new src errors); full unit 765/0 (incl. a new withExperience: false case); integration 63/0.

Deliberately not changed

Knowledge mid-session-refresh TTL stays removed per the "knowledge is persistent" decision.

🤖 Generated with Claude Code

[github-actions]

Explorbot Self-Regression

Commit 3fc32ec · run

Scenario	Result	Attempts	Duration
basic (native)	PASS	1/3	8m
experience: control	OK — failed as expected	1/1	1m
experience: seeded	PASS	1/3	1m

Attempt details

• basic (native) attempt 1 — PASS: login evidence: PASS (post-login plan=true, post-login research=true); research: PASS (files=8, wellFormed=true, keywords=7/3); scenarios: PASS (tests=6/5, features=4/3); tests passed: PASS (5 passed, 0 failed (reporter: 5 passed, 0 failed))
• experience: control attempt 1 — PASS: control: OK — failed as expected (0 passed, 1 failed)
• experience: seeded attempt 1 — PASS: seeded: PASS (1 passed, 0 failed)

---

Session analysis — basic (native):

Session Analysis

The Issues list page functionality was explored: creating issues, searching by text, filtering by status, filtering by label, and clearing filters. All core flows work correctly.

Coverage

• Pages: /issues
• Features: Create issue, search/filter by text, filter by status, filter by label, clear filters

What works

• Create issue — ET-1 Create new issue via toolbar button
• Search by text — ET-2 Search for issue by text and verify filtered results
• Filter by status — ET-3 Filter issues by status and verify results
• Filter by label — ET-4 Filter issues by label and verify results
• Clear filters — ET-5 Clear search filters and verify full list restored