Borrowing it
Nothing to install: this file belongs to florafauna-ai/flora-mcp-skills. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/florafauna-ai/flora-mcp-skills/main/.agents/skills/flora-refine-loop/SKILL.mdgit clone --depth 1 https://github.com/florafauna-ai/flora-mcp-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/florafauna-ai/flora-mcp-skills/flora-refine-loop)<a href="https://agentmods.dev/skills/florafauna-ai/flora-mcp-skills/flora-refine-loop"><img src="https://agentmods.dev/badge/skills/florafauna-ai/flora-mcp-skills/flora-refine-loop/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/florafauna-ai/flora-mcp-skills/flora-refine-loop"><img src="https://agentmods.dev/badge/skills/florafauna-ai/flora-mcp-skills/flora-refine-loop.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00162 | $0.03282 |
| Opus 5 | $0.00081 | $0.01641 |
| Sonnet 5 | $0.00032 | $0.00656 |
| Haiku 4.5 | $0.00016 | $0.00328 |
Grade A, and why
flora-refine-loop scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 301 lines — stays where its author put it; the contents beside it link to each section on GitHub.
flora-refine-loop
What it is
A stopping rule with a generator attached. Anyone can regenerate. The hard part — and the expensive part — is knowing whether the last attempt got closer, and knowing when to stop paying to find out.
The failure this skill exists to prevent has a specific shape. The user doesn't like a result. The agent runs the same prompt again. The model returns a different image that is wrong in a different way. Nobody wrote down what "wrong" meant, so nobody can tell whether round four beat round two. Credits burn at full price while the loop makes no progress, because nothing in it is measuring anything.
That is not iteration. It is a slot machine with a polite interface.
The law
A re-run with an unchanged prompt is not a refinement. It is a re-roll. Every regeneration must name the defect it is fixing and change the prompt to fix it. No named defect, no spend.
The corollary is the part that saves money: if you cannot name what is wrong, you are not ready to spend again. Go and look at the image first. Looking is free.
Looking is free — this is the whole economic argument
FLORA has image-to-text models. Three of them cost nothing:
i2t-gemini-3-5-flash-lite-i2t 0 credits ~4s
i2t-gemini-3-6-flash-i2t 0 credits ~6s
i2t-gemini-3-7-flash-i2t 0 credits ~6s <- default for this skill
So the loop's economics are lopsided in exactly the right direction: evaluating is free, generating is not. Judge every candidate. Regenerate only on a named defect.
Measured on this skill's hand-test: a two-round loop reached the goal for $0.008 —
two generations of t2i-flux-2-klein-4b — with $0.00 spent on the two judgements
that made it converge. Cheap models exist for the generator too; use them while the
prompt is still being tuned and switch to the expensive one only once the loop passes.
If a judge needs to be better than free, i2t-gemini-3-flash and
i2t-claude-sonnet-4-6-vertex-i2t are 8 credits (~$0.008). Do not reach for
i2t-gpt-5-5-i2t (100 credits) or i2t-openai-o3-deep-research (900 credits, 10
minutes) — a judge that costs more than the generation defeats the point.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 12d ago First seen · 301 lines · 162 tokens per session scan A d364215036f6
flora-refine-loop is a skill published in the GitHub repository florafauna-ai/flora-mcp-skills (8 stars, last pushed today), licensed MIT. It adds 162 tokens to every session and 3,282 once invoked, about $0.0008 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
webgl-holographic-foil
A self-contained WebGL2 hero: thin-film interference over a crushed-foil surface whose palette shifts with the viewing angle; move the cursor to tilt the film.
general-video
Author or edit a custom HyperFrames composition when no specialized workflow fits, or when BRIEF.md sets flow: companion. Use for longer or multi-scene pieces, brand and sizzle reels, montages, static loops, static title cards, footage remixes, and freeform builds. Use motion-graphics instead for a short unnarrated…
html-ppt-hermes-cyber-terminal
OpenDesign + BYOK: choosing and wiring your own model, hands-on — cost, quality, and the routing decision. Built as a decision-grade AI literacy deck for engineers, IT, applied-AI teams.
html-ppt-taste-brutalist
16:9 HTML deck in tactical-telemetry / CRT-terminal taste. Deactivated-CRT charcoal slides, white-phosphor monospace, hazard-red accent, scanline overlay, ASCII syntax, density over decoration. Distilled from Leonxlnx/taste-skill brutalist-skill (Tactical Telemetry mode).
chengfeng-check-updates
An environment manager for a video-editing system. It checks whether its skills and runtime—the software needed to run them—are installed and compatible.
diagnostic-stem-delivery
Audio production with diagnostic analysis, timecode parsing from documents, and verified export workflow.