prompt — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited prompt (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
Helps you pick the best prompt engineering technique for your problem, then applies it.
/prompt [query] — recommends technique based on your intent
/prompt [technique] [query] — applies a specific technique
/prompt list — show all 31 techniques
/prompt [category] — list techniques in a category/prompt list:mcp__tachibot-mcp__list_prompt_techniques({ filter: "all" })/prompt [category] (e.g. /prompt creative):mcp__tachibot-mcp__list_prompt_techniques({ filter: "[category]" })Categories: creative, research, research_advanced, analytical, reflective, reasoning, verification, meta, debate, judgment, engineering, structured_coding, decision
/prompt [query] (NO technique specified — RECOMMEND one):This is the key flow. Analyze the user's intent and recommend a technique:
Intent Matching Guide:
| User intent | Best technique | Why |
|---|---|---|
| "Why does X happen?" / understand root cause | first_principles | Strip to atomic truths |
| "How should I build X?" / explore approaches | tree_of_thoughts | Branch 3 paths, evaluate each |
| "What if X?" / speculative | what_if | Wild exploration without limits |
| "Compare X vs Y" / trade-offs | adversarial | Argue FOR then AGAINST |
| "Break down X" / complex problem | decompose | Sub-problems + dependencies |
| "Is X feasible?" / reality check | feasibility | Technical/time/cost/risk |
| "Review X" / check my work | self_consistency | 3 solutions, compare, vote best |
| "Research X" / investigate | investigate | 5W1H analysis |
| "Improve X" / creative ideas | innovate | 3+ unconventional solutions |
| "What am I missing?" / blind spots | alt_view | 5 perspectives: child, scientist, artist, strategist, futurist |
| "Find patterns in X" | patterns | Themes, causality, cycles, anomalies |
| "Synthesize X" / connect findings | integrate | Convergent themes, meta-patterns |
| "Best approach for X?" / need consensus | council_of_experts | Multi-model judge council |
| "Step by step, how?" | chain_of_thought | Identify, breakdown, logic, conclude |
| "Map connections in X" | graph_of_thoughts | Nodes=ideas, edges=connections |
| "Is this safe/ethical?" | constitutional | Critique against principles, revise |
| "Make my prompt better" | meta_prompting | Write better prompt, then solve |
| "Prove/disprove X" | evidence | Support, contradict, cases, stats |
| "What went well/wrong?" | reflect | Patterns, surprises, gaps, next steps |
| "Debate X" / multiple experts | persona_simulation | Simulated expert debate |
| "Apply X idea to Y domain" | creative_use | Cross-domain applications |
| "Analyze X systematically" / structured analysis | analyze | Components → relationships → patterns → conclusions |
| "Solve simplest parts first" / build up | least_to_most | Atomic parts first, build up to hardest |
| "What if this fails?" / risk analysis | pre_mortem | Assume failure, find causes, mitigate |
| "Iteratively fix this code" / self-correct solution | reflexion | Generate→Critique→Revise loop, score 1-10 |
| "Execute multi-step task" / reason-then-act | react | Thought→Action→Observation loops |
| "Explain this code to find the bug" / walkthrough | rubber_duck | Explain line-by-line to novice, flag assumptions |
| "Write tests for X" / test-first approach | test_driven | Edge cases→Tests→Minimal code→Refactor |
| "Plan and implement X from scratch" / new code | scot | Reason in code structures before writing (sequence/branch/loop) |
| "Define the contract for X" / API boundaries | pre_post | State preconditions + postconditions before implementing |
| "Specify behavior for X" / acceptance criteria | bdd_spec | Given/When/Then scenarios before code |
Present 2-3 choices, then judge which is best:
preview_prompt_technique for each)I found 3 techniques that fit. Here's how each would approach your question:
### Option 1: **[technique1]** ([alias])
> [Show the enhanced prompt preview — how it reframes the query]
Strength: [why this angle works]
### Option 2: **[technique2]** ([alias])
> [Show the enhanced prompt preview]
Strength: [why this angle works]
### Option 3: **[technique3]** ([alias])
> [Show the enhanced prompt preview]
Strength: [why this angle works]
**My pick: Option [N] — [technique]**
[1-2 sentences on why this framing will get the best result for your specific question]
Go with this, or prefer another?Wait for user confirmation before executing.
/prompt [technique] [query] (technique specified):Step 1 — Preview (always do this first):
mcp__tachibot-mcp__preview_prompt_technique({
technique: "[technique]",
tool: "grok_reason", // or gemini_brainstorm for creative techniques
query: "[query]"
})Show the user the original query vs enhanced prompt side by side, plus the technique name and target tool.
Step 2 — Confirm: Ask the user if they want to execute, tweak the query, or pick a different technique.
Step 3 — Execute (only after approval):
mcp__tachibot-mcp__execute_prompt_technique({
execution_token: "last" // uses the most recent preview
})Note: Tokens expire after 5 minutes. If expired, re-preview first.
Match the tool to the technique category:
| Category | Best tool | Fallback |
|---|---|---|
| Creative | gemini_brainstorm | grok_brainstorm |
| Research | perplexity_ask | grok_search |
| Analytical | grok_reason | openai_reason |
| Reflective | gemini_analyze_text | kimi_thinking |
| Reasoning | kimi_thinking | openai_reason |
| Verification | openai_reason | grok_reason |
| Debate | grok_reason | openai_reason |
| Judgment | gemini_analyze_text | openai_reason |
| Meta | openai_reason | grok_reason |
| Engineering | qwen_coder | grok_reason |
| Structured Coding | qwen_coder | minimax_code |
| Research Advanced | perplexity_research | kimi_thinking |
| Decision | grok_reason | openai_reason |
/prompt why do users abandon checkout — recommends first_principles/prompt first_principles why do users abandon checkout — applies directly/prompt how should I implement caching — recommends tree_of_thoughts/prompt list — shows all 33 techniques/prompt creative — shows creative techniques only/prompt reflexion fix my authentication middleware — iterate→critique→revise loop/prompt scot implement a rate limiter — reason in code structures first/prompt bdd_spec user registration flow — Given/When/Then before code/prompt pre_mortem deploy to production — assume failure, find causes~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.