How do I install the Hiapi Gpt Image 2 5 skill?

Run npx skillmds@latest add hiapiai/hiapi-gpt-image-2-5 in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Hiapi Gpt Image 2 5 safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Hiapi Gpt Image 2 5?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Hiapi Gpt Image 2 5 free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Hiapi Gpt Image 2 5?

HiAPIAI (@hiapiai) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Hiapi Gpt Image 2 5 by hiapiai — Agent Skill

Name: Hiapi Gpt Image 2 5
Author: HiAPIAI

Hiapi Gpt Image 2 5

Use only for HiAPI GPT Image 2.5 image generation and editing with the exact model IDs gpt-image-2.5-flare or gpt-image-2.5-sunburst. Covers text-to-image, image-to-image, schema validation, dry-run estimates, idempotent async creation, polling, resume, and output download. Do not invoke for GPT Image 2, other GPT Image versions, or unrelated image models.

HiAPIAI Updated Sep 17, 2026

File contents

HiAPI GPT Image 2.5

This released 0.1.0 skill wraps HiAPI's unified async image API for exactly two model IDs:

gpt-image-2.5-flare — the default when no model is selected.
gpt-image-2.5-sunburst — use only when the caller explicitly selects it.

Both IDs expose the same public input contract: one required prompt, optional public reference URLs, an aspect ratio or supported pixel size, quality, background, and output format. Each accepted task produces one image. Do not infer a quality, speed, or visual advantage between Flare and Sunburst from their names.

Required sequence

Decide whether the request is text-to-image (omit image_urls) or image-to-image/editing (send 1–16 accessible HTTP(S) URLs). A local file path is not accepted by this CLI; serve it from an accessible URL or use the direct API with an accessible URL.
Validate the prompt and options locally. Run --dry-run --estimate before a paid create. The estimate reads the current /api/pricing snapshot and creates no task.
Submit once with a stable --idempotency-key. The CLI prints the key and task ID to stderr immediately after acceptance. If acceptance is ambiguous, reuse the same key; never blind-retry with a new key.
Use the default wait, or --no-wait to return after submission. Use --resume-task-id to poll/download an existing task without creating another task. Use --no-save when you only need the output URL.
Keep the downloaded image through technical and creative checks. Treat temporary output URLs as expiring.

Read references/api.md for the field contract, references/workflow.md for delivery and QC, and references/output.md for response handling and recovery.

Safety boundary

--dry-run, --estimate, and local validation are non-billing. Creating a task may spend account balance. This package does not claim a paid generation or client runtime acceptance. Callbacks and HiAPI persistent storage are not implemented as CLI flags in this package; use the direct API snippets in the references when those features are enabled for the account.

---name: hiapi-gpt-image-2-5description: Use only for HiAPI GPT Image 2.5 image generation and editing with the exact model IDs gpt-image-2.5-flare or gpt-image-2.5-sunburst. Covers text-to-image, image-to-image, schema validation, dry-run estimates, idempotent async creation, polling, resume, and output download. Do not invoke for GPT Image 2, other GPT Image versions, or unrelated image models.---# HiAPI GPT Image 2.5This released 0.1.0 skill wraps HiAPI's unified async image API for exactly two model IDs:- `gpt-image-2.5-flare` — the default when no model is selected.- `gpt-image-2.5-sunburst` — use only when the caller explicitly selects it.Both IDs expose the same public input contract: one required prompt, optional public reference URLs, an aspect ratio or supported pixel size, quality, background, and output format. Each accepted task produces one image. Do not infer a quality, speed, or visual advantage between Flare and Sunburst from their names.## Required sequence1. Decide whether the request is text-to-image (omit `image_urls`) or image-to-image/editing (send 1–16 accessible HTTP(S) URLs). A local file path is not accepted by this CLI; serve it from an accessible URL or use the direct API with an accessible URL.2. Validate the prompt and options locally. Run `--dry-run --estimate` before a paid create. The estimate reads the current `/api/pricing` snapshot and creates no task.3. Submit once with a stable `--idempotency-key`. The CLI prints the key and task ID to stderr immediately after acceptance. If acceptance is ambiguous, reuse the same key; never blind-retry with a new key.4. Use the default wait, or `--no-wait` to return after submission. Use `--resume-task-id` to poll/download an existing task without creating another task. Use `--no-save` when you only need the output URL.5. Keep the downloaded image through technical and creative checks. Treat temporary output URLs as expiring.Read [references/api.md](references/api.md) for the field contract, [references/workflow.md](references/workflow.md) for delivery and QC, and [references/output.md](references/output.md) for response handling and recovery.## Safety boundary`--dry-run`, `--estimate`, and local validation are non-billing. Creating a task may spend account balance. This package does not claim a paid generation or client runtime acceptance. Callbacks and HiAPI persistent storage are not implemented as CLI flags in this package; use the direct API snippets in the references when those features are enabled for the account.

Frequently asked questions