chore(deps): update experiments digest to 3e3f4df

[renovate-fullsend]

This PR contains the following updates:

Package	Update	Change
experiments (changelog)	digest	f02b681 → 3e3f4df

---

Warning

Some dependencies could not be looked up. Check the Dependency Dashboard for more information.

---

Configuration

📅 Schedule: (UTC)

• Branch creation
  • At any time (no schedule defined)
• Automerge
  • At any time (no schedule defined)

🚦 Automerge: Disabled by config. Please merge this manually once you are satisfied.

♻ Rebasing: Whenever PR is behind base branch, or you tick the rebase/retry checkbox.

🔕 Ignore: Close this PR and you won't be reminded about this update again.

---

• If you want to rebase/retry this PR, check this box

---

This PR has been generated by Mend Renovate.

[fullsend-ai-review]

🤖 Finished Review · ✅ Success · Started 4:19 AM UTC · Completed 4:24 AM UTC
Commit: d0eee4c · View workflow run →

[github-actions]

Site preview

Preview: https://100cfb5a-site.fullsend-ai.workers.dev

Commit: d0eee4c32a7242a9f03d6cc135550688ee82fbdc

[fullsend-ai-review]

Looks good to me

---

Labels: Renovate bot submodule digest update for experiments

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!

[fullsend-ai-retro]

🤖 Finished Retro · ✅ Success · Started 8:55 AM UTC · Completed 9:05 AM UTC
Commit: d0eee4c · View workflow run →

[fullsend-ai-retro]

Retro: PR #5878 — Renovate submodule digest update

Workflow went well. No novel improvement proposals — all identified opportunities are already tracked in existing issues.

Timeline

Time (UTC)	Event
04:18	Renovate bot creates PR — 1-line experiments submodule pointer bump (f02b681 → 3e3f4df)
04:19	Review agent dispatched (run 30877311258)
04:24	Review agent approves — applied ready-for-merge, dependencies, submodules labels
08:47	Human reviewer rh-hemartin approves
08:47	Added to merge queue
08:53	Merged

What went well

• Correct triage: Agent classified the PR as a mechanical/value-only change and applied trivial scope constraints, so both sub-agents (correctness, style-conventions) made 0 tool calls.
• Correct sub-agent selection: Only dispatched the two always-included sub-agents; correctly skipped security, intent-coherence, docs-currency, and cross-repo-contracts.
• Correct verdict: Approved with no findings, appropriate labels.
• No false positives: Zero unnecessary change requests or comments.

Evidence for existing issues

• pr-review skill: reduce sub-agent dispatch for mechanical/value-only changes like dependency bumps #4060 (reduce sub-agent dispatch for mechanical changes): Review cost $1.37 (720K tokens, Opus) for a 2-line submodule bump. Both sub-agents returned empty findings after 0 tool calls — the orchestration overhead dominates.
• Pre-create FULLSEND_OUTPUT_DIR before launching the agent #3012 (pre-create FULLSEND_OUTPUT_DIR): Agent's first write attempt failed because the output directory didn't exist; recovered with mkdir -p but wasted 1 tool call.
• Skip retro dispatch for bot-authored PRs with zero agent involvement #5295 / Skip retro agent for automated bot dependency PRs #3833 / Skip or fast-path retro stage for automerged bot dependency PRs #3647 (skip retro for bot PRs): This retro run itself adds unnecessary cost on a trivial Renovate PR with no agent code involvement.
• Review agent: use cheaper model for trivial bot-authored PRs #2842 / Add fast-path in review orchestrator for trivial bot-authored PRs #2639 (cheaper model / fast-path for trivial bot PRs): The ~720K token context load for a 2-line diff is disproportionate; a lighter-weight path could reduce cost significantly.