CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

firebase-ai-logic-basics

Integrates Firebase AI Logic (formerly Vertex AI for Firebase) into mobile and web apps (Android, iOS, Web, and Flutter) to call Gemini models from client code. Use when provisioning or initializing Firebase AI Logic, or adding Gemini features to an app: text generation, multimodal input, chat, streaming, structured output, hybrid on-device inference, App Check protection, or switching models with Remote Config.

1.15x
Quality

85%

Does it follow best practices?

Impact

97%

1.15x

Average score across 3 eval scenarios

SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

67%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 well-structured overview skill with an actionable, checkpointed provisioning workflow and a sound one-level-deep reference bundle. The main weaknesses are redundant content (a duplicated model-name warning and TTS detail that belongs in the reference) and several empty capability stubs that give an agent no code and no pointer.

Suggestions

State the 'use current model names' warning once (e.g., in Core Capabilities) instead of repeating it nearly verbatim under Initialization Code References, and drop the duplicated deprecated-model list.

Fill or remove the empty capability stubs: 'Text-Only Generation' and 'Search Grounding' currently have no content — add a one-line description plus a pointer to the matching section of the platform references, or delete the headers.

Trim the TTS capability bullet list in SKILL.md to a two-line summary linking to references/tts_generation.md, and merge the redundant 'References' section into 'Initialization Code References'.

DimensionReasoningScore

Conciseness

The body is mostly efficient (terse CLI commands, pointed warnings), but there is clear tightening opportunity: the 'CRITICAL: Use current model names' warning appears twice nearly verbatim (Core Capabilities and again under Initialization Code References) with the same deprecated-model list; the TTS section inlines ~150 words of capability detail ('24 kHz 16-bit linear PCM (audio/l16)... [Director's Note: ...]') that references/tts_generation.md repeats; and stub headers like 'Text-Only Generation' and 'Search Grounding with the built in googleSearch tool' carry no content. This matches the 3 anchor ('mostly efficient but... could be tightened'); it is not a 2 because there is no conceptual padding (nothing explains what Firebase or Gemini 'is' at length), and not a 4 because the duplicated warning block and TTS detail exceed 'minor instances of over-explanation'.

3 / 5

Actionability

The setup workflow is fully executable — 'npx -y firebase-tools@latest projects:list', 'apps:list', 'init ailogic' — with concrete failure modes ('PERMISSION_DENIED' if not provisioned) and an error-recovery branch ('If you can't run init ailogic yourself... Tell the user to run this command'), and it delegates platform code to real reference files with the instruction to read them 'Before writing code'. It is not a 5 because several capability sections are stubs with no code and no pointer: 'Text-Only Generation' and 'Search Grounding' are empty headers, and 'Generate Images with Nano Banana' states only the billing requirement; not a 3 because the executable commands and verified reference delegation cover the common cases with only minor gaps.

4 / 5

Workflow Clarity

Provisioning is a clear numbered sequence with pre-flight checkpoints ('the active project is marked with "current"', 'Make sure there's at least one Firebase App... registered') and an explicit recovery path when the agent cannot run init ailogic. It is not a 5 because there is no post-action validation after 'init ailogic' (no check that provisioning succeeded) and step 2 simply delegates to per-platform references without sequencing detail; it is above a 3 because checkpoints are explicit rather than 'implicit or missing'.

4 / 5

Progressive Disclosure

SKILL.md is a genuine overview with five one-level-deep reference files (all verified to exist: usage_patterns_android.md, ios_setup.md, usage_patterns_web.md, flutter_setup.md, tts_generation.md) and no nested .md references inside them. It is not a 5 because the same references are listed twice ('Initialization Code References' and a redundant 'References' section), the TTS capability detail is inlined rather than split to its dedicated reference, and capability sections like 'Text-Only Generation' lack a pointer to where their code lives; it comfortably exceeds a 3 because the split itself is appropriate and the references are clearly signaled.

4 / 5

Total

15

/

20

Passed

Description

100%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.

An exemplary description: third-person voice, a clear what, an explicit 'Use when' clause with concrete feature-level triggers, and synonym coverage via the former product name. Every dimension lands on the top anchor without padding or over-claiming.

DimensionReasoningScore

Specificity

Quotes: 'Integrates Firebase AI Logic (formerly Vertex AI for Firebase) into mobile and web apps (Android, iOS, Web, and Flutter) to call Gemini models from client code' plus a concrete capability list: 'text generation, multimodal input, chat, streaming, structured output, hybrid on-device inference, App Check protection, or switching models with Remote Config.' This lists multiple specific concrete actions with comprehensive coverage of the skill's scope, matching the 5 anchor; it exceeds the 4 anchor ('minor gaps in coverage') because every capability named in the body's Core Capabilities section (TTS/image generation excepted) is represented.

5 / 5

Completeness

It explicitly answers 'what' ('Integrates Firebase AI Logic... to call Gemini models from client code') and 'when' ('Use when provisioning or initializing Firebase AI Logic, or adding Gemini features to an app: ...') with concrete trigger phrases, exactly matching the 5 anchor. It is not a 4 because the 'when' clause is already fully explicit and specific rather than 'could be more explicit'.

5 / 5

Trigger Term Quality

Natural trigger phrases include 'provisioning or initializing Firebase AI Logic', 'adding Gemini features to an app', 'call Gemini models', 'chat', 'streaming', 'structured output', and platform names (Android, iOS, Web, Flutter). Synonym coverage is strong — it includes both the current product name and the former name 'Vertex AI for Firebase', which users still say — matching the 5 anchor ('comprehensive coverage of natural terms including synonyms'); no file extensions apply to this domain. It is above the 4 anchor because the terms a user would naturally utter when needing this skill are all present.

5 / 5

Distinctiveness Conflict Risk

The description carves out a clear niche — client-side Gemini integration through Firebase AI Logic — with distinct triggers ('provisioning or initializing Firebase AI Logic', 'App Check protection', 'Remote Config') unlikely to fire for unrelated skills. It disambiguates from generic Vertex AI/GCP skills via 'formerly Vertex AI for Firebase' and 'from client code', matching the 5 anchor ('clear niche with distinct triggers; minimal conflict risk') rather than the 4 anchor, which reserves points for overlap with closely related skills.

5 / 5

Total

20

/

20

Passed

Validation

87%

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

Validation — 14 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

metadata_version

'metadata.version' is missing

Warning

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

14

/

16

Passed

Repository

firebase/agent-skills

Path

skills/firebase-ai-logic-basics/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.