Skip to content
Skills Library
AI & agents · Automation

sol-delegate

/sol-delegate

Use when Sol is the main Codex agent and should delegate execution to Terra, GPT-5.5, GPT-5.4, or another available model while retaining judgment and review.

Licence
MIT
Last repository change
25.10.2026

Assessment and security

90/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content25 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
1
Scripts
0
HTML comments
0
URLs in SKILL.md
0
Deprecation

No signals detected. Repository changed on 25.10.2026, without references to retired models or APIs.

Licence

MIT, open: use and modify while retaining author attribution.

What it does and when to use it

Use when Sol is the main Codex agent and should delegate execution to Terra, GPT-5.5, GPT-5.4, or another available model while retaining judgment and review.

Original author description

Use when Sol is the main Codex agent and should delegate execution to Terra, GPT-5.5, GPT-5.4, or another available model while retaining judgment and review.

How it works

  1. Keep in Sol
  2. Delegate
  3. Write the work order
  4. Review the result
  5. Model routing
Original SKILL.md headings · 51 lines, 450 words
SKILL.md preview
---
name: sol-delegate
description: Use when Sol is the main Codex agent and should delegate execution to Terra, GPT-5.5, GPT-5.4, or another available model while retaining judgment and review.
---

# Sol: Delegate the Doing

Use Sol's scarce frontier-model attention for decisions, not routine execution. Keep ownership of the outcome while routing mechanical work to a cheaper or more appropriate Codex model.

## Keep in Sol

- Understand the user's actual goal and define the work shape.
- Make architecture, scope, risk, and trade-off decisions.
- Write prompts that give delegated work a concrete contract.
- Review delegated output critically against the goal.
- Resolve ambiguity that genuinely requires the main agent's judgment.
- Communicate decisions, evidence, and remaining risks to the user.

## Delegate

Route file searches, routine edits, test runs, mechanical refactors, documentation passes, and bounded research to another available model. Prefer Terra for normal execution, GPT-5.5 or GPT-5.4 when the delegated task needs more reasoning, and Luna when direct model selection becomes available.

When direct Luna selection is unavailable, use Hotline/cmux only if that path is already configured and the user has put it in scope. Do not pretend a model switch occurred merely because the task was described as a Luna assignment.

## Write the work order

Every delegation prompt must state:

1. The outcome to produce.
2. The files or surfaces in scope.
3. Constraints, including files or actions to leave untouched.
4. The checks or evidence required before reporting completion.
5. What the delegated agent must report back.

Delegate independent tasks separately when their outputs can be reviewed independently. Keep dependent work in sequence so the next agent receives the verified result of the previous step.

## Review the result

Treat delegated output as evidence, not authority. Inspect the diff or artifact, run the load-bearing checks, and verify the behavior at the surface the user actually cares about. If the delegated agent found a broader class of problem, follow that diagnosis instead of accepting the narrow patch.

Do not delegate the final judgment about whether the user's goal is met. Do not end with a report that merely says another model changed files; report what changed, what was verified, and what remains uncertain.

## Model routing

Use this default routing unless the task's risk or the user's instruction calls for a different choice:

| Main agent | Execution target | Best fit |
| --- | --- | --- |
| Sol | Terra | Routine edits, searches, tests, and bounded implementation |
| Sol | GPT-5.5 | Complex implementation or debugging needing a stronger delegate |
| Sol | GPT-5.4 | Everyday implementation with moderate reasoning needs |
| Sol | Luna | Fast mechanical work once direct selection exists |

The routing table is a policy, not proof that a model was invoked. Record the actual target and inspect the returned work.

Versions

The database has no other skills named sol-delegate.

Database snapshot: 13.11.2026 · package version 2221ddc79a95 (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 100use-codex
    Delegate independent codebase analysis, adversarial review, debugging, implementation, or patching to OpenAI Codex CLI as an external subagent. Use only when the user explicitly asks to use Codex, requests a Codex review or second opinion, or wants Codex to patch files. Route…
  • 100delegate-codex
    Delegate a self-contained piece of work to a Codex CLI worker (`codex exec`) instead of a Claude subagent, picking the Codex tier by difficulty (luna for what Haiku could do, terra for Sonnet-level work, sol for hard jobs), always on the newest model version available. Use when…
  • 100shellout
    How to delegate a self-contained task to a cheaper pseudo-subagent via `codex exec` (OpenAI models; default gpt-5.6-sol). Use ONLY when the user explicitly invokes /shellout or names codex as the runner — not for general subagent or delegation requests.
  • 100implement
    Delegate a coding task to a Codex subagent (GPT 6.1 Sol by default) that writes files directly in the working directory. Use when the user wants Codex — or "GPT-5.5" — to build, implement, write, add, refactor, fix, or otherwise change code, whether they say "have Codex do X",…
  • 92sol-loop
    Use when the user says "sol loop".
  • 100allocate-agent-models
    Use when delegating OpenAI agent work across GPT-5.6 Sol, Terra, or Luna, especially for new task classes, parallel workers, independent review, model fallback, repeated failures, weak verification, or high-consequence operations.

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.