CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

pika-h3-flex

Generate 3-15 second video with native synchronized audio using Pika H3 Flex through the Pika API (`pika_video`). Use for text-to-video, first/last-frame conditioning, and reference-guided generation from up to 9 images, 3 videos and 3 audio clips, with 480p/768p native output and 1080p/2k/4k upscale. Covers the prompt structure the model responds to, when to turn prompt rewriting off, reference rules, pricing and billing behavior, and resume after a polling timeout.

Quality

90%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

82%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A tight, information-dense skill for the Pika H3 Flex API: concrete parameters, explicit cost/validation behavior, and thorough error recovery with no padding. The main gaps are the absence of a worked example call or prompt and no consolidated end-to-end workflow narrative.

Suggestions

Add one fully worked example — a sample `pika_video` request plus the filled-in shot-list prompt it would use — so the five-part prompting structure has a concrete model to copy.

Consolidate the workflow into an explicit ordered sequence (draft at 480p → `dry_run` preflight → generate → check `data.cost_status` is `settled` → rerun approved take at delivery resolution) instead of distributing it across sections.

Consider moving the full per-resolution price tables and detailed reference rules into a `references/` file, keeping SKILL.md as a leaner overview with cost decision rules inline.

DimensionReasoningScore

Conciseness

The body is dense with Pika-specific operational facts (parameter enums, frame-grid rounding, per-resolution pricing, error codes) and never explains concepts Claude already knows; e.g. 'Output rounds up to the model's frame grid (124 frames for a 5 second request at 24 fps)' and 'honor `data.retry_after`, then submit again'. Every section carries non-inferable API detail, matching the 'every token earns its place' anchor.

5 / 5

Actionability

Parameter names, enums, defaults, and constraints are fully specified ('Set `dry_run: true` to validate the request', 'Call `pika_video` again with that object (and a longer `poll_timeout`)'), and the five-part prompt structure is concrete. It falls short of anchor 5 only because there is no worked example — no sample `pika_video` invocation or filled-in prompt showing the shot-list format in practice.

4 / 5

Workflow Clarity

Validation checkpoints and feedback loops are present: `dry_run` preflight with quoted price, per-error recovery actions (`insufficient_balance`, `rate_limited` with `retry_after`, `resume_job` on poll timeout), and a draft-at-480p-then-rerun-at-delivery-resolution path. It sits below anchor 5 because these steps are distributed across sections rather than consolidated into one explicit end-to-end sequence.

4 / 5

Progressive Disclosure

The single file (~108 lines) is well-organized with clear sections (Route and billing, Operations, Reference rules, Shared parameters, Prompting, Failure handling, Provenance) and no nested references. It does not reach anchor 5 because the body exceeds the under-50-lines simple-skill case and the full pricing/rate tables and prompting guide could arguably live in reference files, though nothing is misfiled or buried.

4 / 5

Total

17

/

20

Passed

Description

92%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong description: it states exactly what the skill does with concrete numbers, includes an explicit 'Use for' trigger clause covering all operation modes, and is tightly scoped to the Pika provider so conflict risk is minimal. The only weakness is slightly jargon-leaning trigger vocabulary that misses a few natural user phrasings.

DimensionReasoningScore

Specificity

Multiple concrete actions with hard numbers are enumerated: 'Generate 3-15 second video with native synchronized audio', 'text-to-video, first/last-frame conditioning, and reference-guided generation from up to 9 images, 3 videos and 3 audio clips, with 480p/768p native output and 1080p/2k/4k upscale'. This comprehensively covers the tool's operation modes and outputs, matching anchor 5 rather than the minor-gaps level of 4.

5 / 5

Completeness

The 'what' is explicit ('Generate 3-15 second video with native synchronized audio using Pika H3 Flex through the Pika API (`pika_video`)') and the 'when' is an explicit trigger clause ('Use for text-to-video, first/last-frame conditioning, and reference-guided generation from up to 9 images...'). Both are answered with concrete trigger cases, matching anchor 5.

5 / 5

Trigger Term Quality

Good natural terms are present — 'video', 'text-to-video', 'audio', 'Pika' (via the named API), 'upscale' — which a user needing this skill would plausibly say. It misses anchor 5 because common natural synonyms are absent: no 'clip', 'footage', 'animate an image', or plain 'make a video from images' phrasing, and 'first/last-frame conditioning' is jargon rather than a user phrase.

4 / 5

Distinctiveness Conflict Risk

The description is scoped to a named provider, model, and tool ('Pika H3 Flex', 'Pika API', '`pika_video`'), giving it a clear niche with distinct triggers and minimal risk of firing for a generic video-generation request aimed at another tool.

5 / 5

Total

19

/

20

Passed

Validation

100%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository

calesthio/OpenMontage

Path

.agents/skills/pika-h3-flex/SKILL.md

Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.