feat: add trtc-ai-service skill for Conversational AI (AI customer service)#8
Open
Burgerham4R wants to merge 8 commits into
Open
feat: add trtc-ai-service skill for Conversational AI (AI customer service)#8Burgerham4R wants to merge 8 commits into
Burgerham4R wants to merge 8 commits into
Conversation
added 8 commits
June 22, 2026 17:16
- Add skills/trtc-ai-service/ with 6 capability modules, 2 scenario recipes, 12 tech-stack adapters, and automation scripts - Add AI Service fast-path in skills/trtc/SKILL.md to route AI customer service requests directly, bypassing standard onboarding/topic pre-flight checks - Register trtc-ai-service in CLAUDE.md product list and routing logic - Register trtc-ai-service in CODEBUDDY.md product table and routing rules - Zero changes to trtc-topic/SKILL.md (Plan A: independent root skill, zero intrusion)
….md routing Root cause: installer used hardcoded SKILL_NAMES array that missed trtc-ai-service, and AGENTS.md only referenced CLAUDE.md without being self-contained for Codex. Changes: - bin/cli.js: Replace hardcoded SKILL_NAMES with dynamic getSkillNames() that auto-discovers skill dirs containing SKILL.md. trtc is always first, rest sorted alphabetically. Also dynamic listSkills() output. - ai-instructions/base.md: Add AI Service routing rules so generated AGENTS.md / CODEBUDDY.md are self-contained (Codex only installs AGENTS.md) - AGENTS.md: Sync AI Service routing rules from base.md source - README.md / README.zh.md: Add AI Service product row with Web check This ensures any new skill directory added to skills/ is automatically discovered and installed without manual code changes.
…nversational AI 1. Sync code updates from upstream ai-customer-service-skill: - capabilities/conversation-core: server.py, trtc_client.py - capabilities/human-handoff: models.py, service.py, router.py, summary_link.py - capabilities/human-handoff: NEW feedback_store.py - capabilities/session-summary: router.py, summarizer.py - scenarios/customer-service: recipe.yaml, README, all UI files (admin-board + voice-customer-service) - start.sh, README.md, README.zh-CN.md 2. Rename product 'AI Service' → 'Conversational AI' across all files: - README.md / README.zh.md: product table + new Conversational AI scenarios section - CLAUDE.md: product identification list - CODEBUDDY.md: product table - ai-instructions/base.md + AGENTS.md: routing section title and description - skills/trtc/SKILL.md: fast-path section title and triggers Customer Service is now correctly positioned as a scenario of Conversational AI, not a product itself.
…scenarios - AI Customer Service: available — describe in plain language - AI Oral Practice: coming soon — describe in plain language Remove skill-internal terminology (customer-service/custom-builder) from README
…rvice SKILL.md The 114-line wrapper was missing critical execution details: - Three-keys configuration (Tencent Cloud API, TRTC SDK, LLM API) - Step-by-step beginner guidance (Path A / Path B) - Environment checks, error recovery, tool whitelist - Security constraints, design standards Now uses the complete 900+ line SOP from the upstream ai-customer-service-skill repo, with official frontmatter and agent-skills entry points section preserved.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Add a new
trtc-ai-serviceskill that enables users to build AI-powered customer service applications using TRTC Conversational AI — from zero to a working demo, or integrate conversational AI capabilities into an existing project.Approach: Independent Root Skill (Zero Intrusion)
The AI customer service skill is added as an independent root skill under
skills/trtc-ai-service/. A new "Conversational AI fast-path" section inskills/trtc/SKILL.mdroutes AI customer service requests directly, completely bypassingtrtc-topic's Pre-flight check (which only covers Conference Web scenarios in v1).Zero changes to
trtc-topic/SKILL.md— the defense-in-depth gate remains untouched.Changes
Modified files (8)
skills/trtc/SKILL.mdCLAUDE.mdCODEBUDDY.mdai-instructions/base.mdAGENTS.mdbin/cli.jsSKILL_NAMESwith dynamicgetSkillNames()— auto-discovers skill dirs containingSKILL.md,trtcsorted firstREADME.mdREADME.zh.mdNew files (~174, all under
skills/trtc-ai-service/)SKILL.md— Full execution SOP (931 lines): three-key configuration, Path A (Quick Start demo) / Path B (Integrate into My System), error recovery, tool whitelistcapabilities/— 6 atomic capability modules: conversation-core, knowledge-base, human-handoff, tool-calling, session-summary, digital-humanscenarios/— 2 scenario recipes: customer-service (demo UI), custom-builder (integration wizard)auto_adapters/— 4-class 12-tech-stack adapters: frontend-spa (React/Vue/Angular), java-backend (Spring/Quarkus), node-backend (Express/Fastify/Koa), python-backend (Flask/FastAPI/Django)scripts/— Automation: credential verification, capability assembly, post-install patch, contract adaptationtests/— 10 test filesstart.sh— Bootstrap scriptNot modified
skills/trtc-topic/SKILL.md— 0 lines changed (zero intrusion)skills/trtc-onboarding/reference/supported-matrix.md— not needed (AI service bypasses onboarding)knowledge-base/index.yaml— not needed (independent capability model)Conversational AI — Supported Scenarios
Testing
npxinstall — all 7 skills (includingtrtc-ai-service) correctly installed to.codex/skills/trtc-ai-service/SKILL.mdWhy dynamic skill discovery in
bin/cli.jsThe installer previously used a hardcoded
SKILL_NAMESarray. This PR replaces it withgetSkillNames()that dynamically scansskills/for directories containingSKILL.md. This ensures future skill additions are automatically discovered and installed without manual code changes.