Executable Skill & Tool Library Crystallization (Voyager-Style Verified Code Reuse)
Turn multi-step trajectories that succeeded in an isolated sandbox into parameterized, unit-tested, reusable skills (`SKILL.md` + deterministic scripts) so future agent runs invoke a verified tool instead of re-deriving brittle multi-turn plans.
At a glance — Oct 2026
Updated 2026-10-28 · forward this page: /jtbds/skill-library-crystallization#briefPromotion class: deterministic_source_reverified · 8.7 ± 0.5 (8.2–9.2 · editorial_prior) · Pricing verified 2026-10-28.
- Depot Pick
- E2B · 8.65/10 for this job (curated estimate, high confidence)
E2B boots isolated Firecracker microVMs in ~150ms ($0.000014/vCPU-s, $0.0000045/GiB-s as of 2026-10-28), providing the clean-room execution substrate required to synthesize a candidate script, run unit tests + adversarial edge cases, and verify zero side effects before crystallizing it into the agent's skill library.
- Runner-up
- Letta Cloud · 7.46/10
- When not to use the pick
- Need stateful, persistent agent memory blocks attached to an agent runtime — Letta; need the evaluation harness to grade candidate skills across 100+ test inputs before promotion — Braintrust; need full git-workspace container persistence across days — Daytona.
- Lock-in / migration
- The pick has lower stored lock-in risk than runner-up Letta Cloud (4 vs 5.5).
- Sourced price (no comparable unit)
Pick: no sourced price in this job's unit (no_matching_sourced_price).
No cheapest SKU (no_comparable_unit_for_job). No price was invented.
- Free option (runnable / substitutable)
- No runnable free tier lists this job. A cheap paid SKU or benchmark dataset is not a runnable free option.
- Failure modes to watch
- No failure mode plainly applies to this job (no_failure_mode_plainly_applies).
Cite: Agent Provider Depot, Depot Pick for skill-library-crystallization: e2b (editorial prior, scores as of 2026-10-28). Agents: GET /api/decide?jtbd=skill-library-crystallization
E2B
E2B boots isolated Firecracker microVMs in ~150ms ($0.000014/vCPU-s, $0.0000045/GiB-s as of 2026-10-28), providing the clean-room execution substrate required to synthesize a candidate script, run unit tests + adversarial edge cases, and verify zero side effects before crystallizing it into the agent's skill library.
Pricing (directional): Firecracker microVM sandboxes billed per second; templates; code interpreter for agents.
Lock-in risk: higher = more lock-in (worse). Overall already inverts this dimension.
Depot default code sandbox for agents. Use for untrusted code; Daytona when you need persistent devworkspaces. Also the depot default Linux desktop sandbox for computer-use (e2b-desktop + VNC stream) when you already standardize on E2B.
Quote for Executable Skill & Tool Library Crystallization (Voyager-Style Verified Code Reuse)
Depot Pick stays E2B. Prices below are sourced SKU rows only (as of 2026-10-28). Input and output are not blended. GET /api/quote?jtbd=skill-library-crystallization
Nothing was compared. no comparable sourced unit for this job; chat input $/1M and other units were not borrowed pickIsPriceWinner is null.
- 1 GiB memory usd_per_gib_second $0.0000045 source as of 2026-10-28
- 1 vCPU usd_per_vcpu_second $0.000014 source as of 2026-10-28
- 2 GiB memory usd_per_sandbox_second_ram_charge $0.000009 source as of 2026-10-28
- 2 vCPU usd_per_sandbox_second_vcpu_charge $0.000028 source as of 2026-10-28
- 4 GiB memory usd_per_sandbox_second_ram_charge $0.000018 source as of 2026-10-28
- 4 vCPU usd_per_sandbox_second_vcpu_charge $0.000056 source as of 2026-10-28
2 of 4 candidates have no sourced price row. 4 are excluded from the Matching price comparison (none). no comparable sourced unit for this job; chat input $/1M and other units were not borrowed
Runners-up
Letta Cloud
Choose when the agent IS a long-lived stateful runtime. For bolt-on memory use Mem0/Zep.
Braintrust
Depot pick when evals are the product quality gate. Pair with Helicone/Langfuse for cheap request logs if…
Daytona
Choose when agents need sticky filesystems and full dev environments. E2B for disposable code exec. Confirm…
When NOT to use the top pick
Need stateful, persistent agent memory blocks attached to an agent runtime — Letta; need the evaluation harness to grade candidate skills across 100+ test inputs before promotion — Braintrust; need full git-workspace container persistence across days — Daytona.
Full ranking (score table)
Ranked with this JTBD's weights. Scores are curated estimates unless marked otherwise. Agents: prefer /api/decide then this table for tie-breaks.
| # | Provider | Overall | Fit | $/perf | Latency | DX | Lock-In Risk | As of |
|---|---|---|---|---|---|---|---|---|
| 1 | E2BDepot Pick | 8.7 | 9.4 | 8.0 | 8.6 | 9.4 | 4.0 | 2026-10-28 |
| 2 | Braintrust | 8.3 | 9.2 | 7.5 | 8.6 | 8.8 | 5.0 | 2026-10-28 |
| 3 | Daytona | 8.3 | 8.8 | 8.2 | 8.2 | 8.8 | 3.0 | 2026-10-28 |
| 4 | Letta Cloud | 7.5 | 8.0 | 7.8 | 7.2 | 8.0 | 5.5 | 2026-10-28 |
Score methodology for this job
Inspired by Voyager (Wang et al.): an agent only promotes a synthesized tool/skill into its reusable library after the code passes automated unit tests and exit-code verification inside a disposable sandbox, preventing broken helpers from polluting future prompts.
capabilityFit: 28%pricePerformance: 16%latencyReliability: 18%developerDx: 16%privacyPosture: 10%lockInRisk: 6%maturity: 6%