research

Compute a confidence interval for a mean (normal approximation). FREE.

Typical input {"mean": 72.4, "std_dev": 8.1, "n": 64, "confidence_pct": 95} returns {"mean": 72.4, "margin_of_error": 1.9845, "interval": [70.4155, 74.3845], "note": "..."}. Errors: on invalid, missing, or malformed input this tool never raises a protocol error — it returns {"error": ""} (for example {"error": "confidence 90/95/99, n>=2, std_dev>=0"}). Every call is read-only and idempotent, so after correcting the input it is always safe to retry.

Format a citation in APA 7, MLA 9, or Chicago author-date style. FREE.

Typical input {"style": "apa", "authors": ["Curie, Marie"], "year": 1911, "title": "Radium and radioactivity", "container": "Century Magazine"} returns {"style": "apa", "citation": "Curie, M. (1911). Radium and radioactivity. Century Magazine.", "note": "..."}. Errors: on invalid, missing, or malformed input this tool never raises a protocol error — it returns {"error": ""} (for example {"error": "style must be apa, mla, or chicago"}). Every call is read-only and idempotent, so after correcting the input it is always safe to retry.

Load a product's free gateway skill with its complete instructions. FREE.

Typical input {"slug": "thesis-advisor"} returns {"slug": ..., "skill": "", "instructions": ""}. Errors: on invalid, missing, or malformed input this tool never raises a protocol error — it returns {"error": ""} (for example {"error": "unknown slug ''. Use list_products."}). Every call is read-only and idempotent, so after correcting the input it is always safe to retry.

Load one product in full: its persona plus every paid skill. PREMIUM (license).

Typical input {"slug": "thesis-advisor"} returns {"slug": ..., "name": ..., "persona": "", "skills": [{"name": ..., "instructions": ...}, ...], "free_skill": {...}}. Errors: on invalid, missing, or malformed input this tool never raises a protocol error — it returns {"error": ""} (for example {"error": "unknown slug ''"}). Every call is read-only and idempotent, so after correcting the input it is always safe to retry.

Load one paid skill's complete instructions from a product. PREMIUM (license).

Typical input {"slug": "thesis-advisor", "skill_name": "Outline Builder"} returns {"slug": ..., "skill": ..., "instructions": ""}. Errors: on invalid, missing, or malformed input this tool never raises a protocol error — it returns {"error": ""} (for example {"error": "unknown slug ''"}). Every call is read-only and idempotent, so after correcting the input it is always safe to retry.

List every product in the Research Desk line with its included skills. FREE.

Takes no arguments. Returns a list of 7 product objects, each {"slug": "thesis-advisor", "name": ..., "tagline": ..., "skills": ["Skill A", ...], "free_skill": "Gateway Skill Name"}. Use the returned slug values with get_free_skill, get_full_product, or get_full_skill. Errors: on invalid, missing, or malformed input this tool never raises a protocol error — it returns {"error": ""}. Every call is read-only and idempotent, so after correcting the input it is always safe to retry.

Calculate the survey sample size needed for a confidence level and margin. FREE.

Uses maximum variance (p=0.5) with a finite-population correction when population is given. Typical input {"population": 5000, "confidence_pct": 95, "margin_pct": 5} returns {"required_sample": 357, "assumptions": "p=0.5 (max variance), random sampling"}. Errors: on invalid, missing, or malformed input this tool never raises a protocol error — it returns {"error": ""} (for example {"error": "confidence_pct must be 90, 95, or 99"}). Every call is read-only and idempotent, so after correcting the input it is always safe to retry.

Describe a numeric dataset: center, spread, quartiles, and outliers. FREE.

Typical input {"numbers": [12, 15, 14, 90, 13]} returns {"n": 5, "mean": 28.8, "median": 14.0, "std_dev": ..., "min": 12, "max": 90, "q1": ..., "q3": ..., "iqr_outliers": [90], "skew": "right (mean > median)"}. Errors: on invalid, missing, or malformed input this tool never raises a protocol error — it returns {"error": ""} (for example {"error": "no numbers"}). Every call is read-only and idempotent, so after correcting the input it is always safe to retry.