Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Get Started Free →

Train Kimi for Smithers · smithersai | DecimalAI

ai mlby smithersai

Train Kimi for Smithers

smithersai/train-kimi-smithers

Source smithersai/smithers ↗License MIT

Prepare, launch, and evaluate Kimi LoRA fine-tuning for Smithers TypeScript/JSX workflow authoring.

SkillScore95Maintained

Automatic

AI Engineer pack

Last verified 8/21/2026

What it changes

Passed 19 of 20 test cases with the skill, and 4 of 20 without it.

Without skill18%

With skill95%

Checks it passed that the model alone missed

Sets LoRA rank to 8
Configures the SFT run for 1 epoch
Sets SFT parameters to 1 epoch and LoRA rank 8

and 22 more in the full report

17 fixed4 passed either way1 still failing0 broke

Pass-rate lift: +77 pp
Tokens: -10%
Agent turns: 0%
Response time: -58%

Model gemini-3.6-flashMeasured 8/21/2026v121 of 22 cases passed with skill · 20 comparable

Tokens, turns and time describe resource use, not a dollar price.

Use this skill

Save it into your project's skills folder. Claude Code picks it up in its next session.

curl -fsSL --create-dirs -o .claude/skills/smithersai-train-kimi-smithers/SKILL.md https://app.decimal.ai/s/smithersai-train-kimi-smithers/SKILL.md

Want it pinned and updatable? pip install "decimalai>=0.15", then decimalai skills add smithersai-train-kimi-smithers, and start Claude Code with claude --plugin-dir "$(git rev-parse --show-toplevel)/.agents"

1 · One-time — install the CLI (docs ↗)

pip install "decimalai>=0.15"

2 · Get the skill — writes .agents/skills/smithersai-train-kimi-smithers/ and pins it in .decimal/skills.lock

decimalai skills add smithersai-train-kimi-smithers

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/smithersai-train-kimi-smithers/SKILL.md https://app.decimal.ai/s/smithersai-train-kimi-smithers/SKILL.md

Raw file: /s/smithersai-train-kimi-smithers/SKILL.md ↗ · machine-readable summary at /s/smithersai-train-kimi-smithers.json ↗

Safety checks: 3 of 3 passed

Code scan CleanIntent PassedContent Clean

See it in action

One of 15 test cases the skill fixed. Same model, same prompt, with and without the skill.

1 of 15

Prompt

We want to deploy an existing fine-tuned model `accounts/my-org/models/my-adapter` for production inference on Fireworks AI using the official Fireworks CLI tool. What exact subcommand creates the deployment?

What was checked, without and with the skill
Checked	Without skill	With skill
Executes fctl create deployment to create the model deployment	Missed	Passed

case-20 · gemini-3.6-flash · 8/21/2026 · answers are capped and redacted

What's inside

The instructions your model reads when the skill loads.

SKILL.md84 lines12 filesincludes scripts

Train Kimi for Smithers

Build the corpus locally, establish a base-model score, run the smallest useful LoRA SFT, and promote it only when the held-out benchmark improves.

Start here

Run from the Smithers repository root:

node skills/train-kimi-smithers/scripts/prepare-dataset.mjs \
  --out artifacts/kimi-smithers
node skills/train-kimi-smithers/scripts/validate-dataset.mjs \
  artifacts/kimi-smithers/train.jsonl \
  artifacts/kimi-smithers/validation.jsonl

The builder uses only canonical workflow files listed in assets/source-manifest.json. It splits by source file, never by generated prompt variant, so validation cannot contain another wording of a training completion. Read methodology.md before changing sources, variants, or the system prompt.

Choose the model path

Prefer Kimi K2.7 Code managed LoRA SFT for a self-serve run now. Its Fireworks model is publicly marked tunable and coding-focused.
Prefer Kimi K3 Serverless Training after Fireworks enables the private preview on the account. K3 is open-weight, but its public model card currently says fine-tuning is unavailable outside the preview.
Do not fall back to K2.5: K2.7 is open-weight and self-serve tunable. Use K2.5 only if the live K2.7 model or training shape is temporarily unavailable.

Read fireworks.md completely before account setup, upload, job creation, promotion, or deployment. Recheck every linked live model, price, and CLI page because provider support changes quickly.

Establish the baseline

After the account has an API key, score the public base model before training:

Track record

Benchmark

19 / 20

test cases passed, vs 4 / 20 without

Adoption

Users: 0
Times used: 0

Safety and source details

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/3/2026

View 15 scanner checks

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

smithersai / smithers

skills/train-kimi-smithers/SKILL.md

View on GitHub ↗

License MITImported 8/3/2026Version v1Bundled 11 files · scripts

More skills from the same source and nearby in the registry, with their headline benchmark lift.

Ready to supercharge your AI agents?

See it in action

What's inside

Train Kimi for Smithers

Start here

Choose the model path

Establish the baseline

Track record

Benchmark

Adoption

SkillSafety

Source

Related skills