Designing conversational flows for website chatbots and AI agents. Intent recognition architecture, branching logic, fallback handling, escalation to human, conversation analytics. Honest about scripted-bot (rigid trees, fail edge cases), hallucinating-bot (LLM without structure, makes things up), and structured-guided-conversation (LLM-powered with intent architecture and fallback discipline) patterns. Distinguishes chatbot DESIGN (this skill) from chatbot IMPLEMENTATION (engineering and platform work). Triggers on chatbot, conversational AI, AI agent, chat widget, intent design, conversational flow, bot escalation, LLM grounding. Also triggers when a chatbot is hallucinating, when a scripted bot is failing edge cases, or when a chatbot is being scoped for the first time.
53
60%
Does it follow best practices?
Impact
—
No eval scenarios have been run
Passed
No known issues
Optimize this skill with Tessl
npx tessl skill review --optimize ./skills/chatbot-flow-design/SKILL.mdQuality
Discovery
100%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This is an excellent skill description that is comprehensive, specific, and well-structured. It clearly defines its scope (design vs. implementation), lists concrete capabilities, provides rich natural trigger terms including scenario-based triggers, and explicitly delineates its boundaries. The description is detailed without being padded, and uses appropriate third-person voice throughout.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple specific concrete actions and concepts: conversational flow design, intent recognition architecture, branching logic, fallback handling, escalation to human, conversation analytics. Also explicitly distinguishes between three design patterns (scripted-bot, hallucinating-bot, structured-guided-conversation) with concrete characterizations of each. | 3 / 3 |
Completeness | Clearly answers both 'what' (designing conversational flows, intent recognition, branching logic, fallback handling, etc.) and 'when' with explicit trigger terms and scenarios ('Triggers on chatbot, conversational AI...', 'Also triggers when a chatbot is hallucinating...'). The 'when' guidance is thorough and explicit. | 3 / 3 |
Trigger Term Quality | Excellent coverage of natural terms users would say: 'chatbot', 'conversational AI', 'AI agent', 'chat widget', 'intent design', 'conversational flow', 'bot escalation', 'LLM grounding'. Also includes scenario-based triggers like 'chatbot is hallucinating' and 'scripted bot is failing edge cases' which are very natural user phrasings. | 3 / 3 |
Distinctiveness Conflict Risk | Explicitly distinguishes chatbot DESIGN from chatbot IMPLEMENTATION, which is a strong boundary-setting move that reduces conflict with engineering/platform skills. The specific focus on conversational flow architecture and intent design creates a clear niche distinct from general AI development or coding skills. | 3 / 3 |
Total | 12 / 12 Passed |
Implementation
20%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This skill reads as a well-structured essay on chatbot design philosophy but fails as an actionable skill for Claude. It is extremely verbose, repeating core concepts multiple times, and provides zero concrete artifacts — no templates, schemas, sample flows, example outputs, or executable guidance. The 12-consideration framework and reference file structure show good organizational thinking, but the content needs to be dramatically condensed and made actionable.
Suggestions
Cut content by 50-60%: remove the closing section (restates everything), reduce the 'what this skill covers' cross-referencing, eliminate repeated explanations of the scripted/hallucinating/structured distinction, and trust Claude to understand concepts like RAG and fallback without explanation.
Add concrete, actionable artifacts: include a sample intent map schema (JSON or YAML), a fallback decision tree template, a sample conversation flow for a common use case, and an example escalation handoff format that Claude can actually produce when asked to design a chatbot.
Add a workflow with validation: replace the 12-consideration list with a step-by-step design process that includes checkpoints like 'verify intent coverage against conversation logs' or 'test fallback paths with 5 out-of-scope queries before finalizing.'
Provide the referenced bundle files or remove references to them — currently 9 files are referenced but none exist, which means the progressive disclosure structure is hollow.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Extremely verbose at ~300+ lines. Extensive philosophical framing, repeated restatements of the same concepts (the scripted/hallucinating/structured distinction is explained at least 3 times), explanations of concepts Claude already knows (what RAG is, what fallback means), and a closing section that restates everything. The 'what this skill covers' section spends significant tokens on cross-referencing other skills. Much of this could be cut by 60%+ without losing actionable content. | 1 / 3 |
Actionability | No concrete code, commands, templates, or executable examples anywhere. The content describes concepts and principles at a high level but never provides a concrete intent map schema, a sample conversation flow diagram format, a fallback decision tree template, or any artifact Claude could actually produce. It reads as an essay about chatbot design philosophy rather than actionable instructions for designing one. | 1 / 3 |
Workflow Clarity | The 12-consideration framework provides a reasonable checklist for auditing or designing a chatbot, and the sections are logically sequenced. However, there are no validation checkpoints, no explicit 'do X then verify Y' steps, and no feedback loops. The workflow is more of a conceptual outline than a step-by-step process with verification gates. | 2 / 3 |
Progressive Disclosure | References to 9 separate reference files are well-organized and clearly signaled at one level deep. However, no bundle files are provided, so these references point to non-existent content. The main SKILL.md itself is monolithic and verbose — much of the inline content repeats what should be in the reference files, undermining the progressive disclosure structure. | 2 / 3 |
Total | 6 / 12 Passed |
Validation
90%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 10 / 11 Passed
Validation for skill structure
| Criteria | Description | Result |
|---|---|---|
frontmatter_unknown_keys | Unknown frontmatter key(s) found; consider removing or moving to metadata | Warning |
Total | 10 / 11 Passed | |
8e70d03
Table of Contents
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.