17 — Adoption Roadmap: Day 1 / Week 1 / Month 1

Summary

Sequenced for one operator; the governing split is automatic beats disciplined — anything that depends on remembering loses to anything that is a default. Each item names its automation vehicle (hook / skill / plugin / settings / jackin❯) or is flagged DISCIPLINE with its decay risk.

• Day 1 is configuration, not behavior: fix the double-registered hooks, pin subagent models, add two CLAUDE.md guard-lines, stop mid-session model switches. ~1 hour, no quality risk.
• Week 1 instruments before it optimizes: OTel token dashboard + the harness screening run, then effort tiering and register choice with canaries, because those two carry the quality risk.
• Month 1 moves the savings into infrastructure: context editing via SDK paths, routing flips validated per task class, and the jackin❯ token-pack so every launched container is born optimized — discipline converted to defaults at launch.rs / RoleManifest / CapsuleConfig.
• Standing rule: re-audit after every Claude Code release — the fixed prefix once grew ~70k tokens in 5 days (GH #45188); platform drift dwarfs micro-optimizations.

Question and scope

Which validated techniques can be adopted immediately, which require experiments, and which remain research-only?

Method

The adoption order ranks techniques by evidence strength, reversibility, dependency, and validation cost, separating direct operating guidance from experiments and research-only ideas.

Findings

The evidence orders adoption by automation and validation risk: establish safe defaults first, measure a baseline before changing behavior, and move validated policies into infrastructure only after they pass the harness.

Implications for jackin❯

Day 1 (≈1 hour, all CLAUDE-CODE-TODAY, zero quality risk)

Explicitly not Day 1: register compression beyond what you already run, effort below high, any context-clearing automation — those need Week 1's instrumentation first.

Week 1 (instrument → screen → tier)

1. Stand up measurement (01 §4): CLAUDE_CODE_ENABLE_TELEMETRY=1, OTLP → any collector, panel on claude_code.token.usage by type and agent.name, plus claude_code.cost.usage. Without this, every later claim about your own savings is folklore. (Hook/infra; hours.)
2. Run the harness screening arm (16): n=12 tasks + 6 canaries on your current setup — this is the baseline every later change is judged against. (Script; ~an evening of wall-clock, mostly unattended.)
3. Effort tiering with canaries (09 §1): /effort medium for routine work, high for hard tasks; re-run canaries; record the thinking-share delta from your dashboard (closes the biggest open measurement in the dossier). DISCIPLINE until month 1 bakes per-task defaults.
4. Register decision (04/02): caveman-ultra ≈58.5% of visible prose at real caveat risk — keep, or swap to plain terse-instruction (09 §3) which captures most of the win with less misread risk. Decide on canary results, not vibes. Skip wenyan: no token advantage over ultra (02 §5).
5. Structured-output sidecars (09 §5): any scripted extraction/classification moves to schema-constrained SDK calls — one stable schema per thread (cache pitfall).

Month 1 (SDK + infrastructure tier)

1. Context editing on long-running SDK/headless flows (12/08): clear_tool_uses_20250919 with clear_at_least sized above the re-write cost; assert via applied_edits and cache fields. Claude Code interactive already auto-compacts; don't double-manage.
2. Routing flip, validated: advisor pattern first (T1-backed, NEGATIVE-COST), then Sonnet-main + escalation per task class that passes n=30 confirmation (15 §3 — the 5–6.6x lives or dies here).
3. Batch lane for nightly sweeps (audits, doc regen, triage): Batch API + caching stack (reads at 0.05x effective; 07/12).
4. The jackin❯ token-pack — discipline becomes infrastructure (14 §jackin❯; Explore-mapped insertion points, this repo):

• env defaults into the container env assembly at crates/jackin-runtime/src/runtime/launch.rs:590-717 (effort tier, telemetry on, attribution header policy);
• [token_policy] block in jackin.role.toml (RoleManifest, crates/jackin-core/src/manifest.rs) → per-role model/effort/register defaults, hooks installed via the existing runtime_setup.rs symlink pattern;
• CapsuleConfig (crates/jackin-protocol/src/lib.rs:25-45) → /jackin/run/agent.toml → build_agent_command so every spawned agent gets model/effort flags without operator action;
• OTLP plumbing already propagates (launch.rs:694-723) — extend the existing traces with the token dashboard so each launched capsule reports its class-split spend.
• Add the compression-hygiene linter (14): CI fails when always-loaded context (root AGENTS.md chain, role files) grows past a token budget — measured with the free count_tokens endpoint in CI.

5. Re-audit cadence: pin a monthly (and post-release) re-run of 02's baseline audit script — prefix size, MCP schema mass, hook payloads. Platform drift is the biggest unmanaged variable (06 §TL;DR, GH #45188).

Automatic vs disciplined — the full split

For jackin❯, the adoption roadmap: day 1 / week 1 / month 1 evidence identifies which mechanisms are safe to adopt directly and which still require workload-specific validation.

Limitations and unknowns

Quantitative effects are modeled from the dossier baseline and remain workload-specific. Effort and register changes require the stated canary and harness gates before becoming defaults.

Sources

Related work

• Token-optimization techniques dossier
• Validation harness
• Context engine