Loading ToolBench
Agentic revenue intelligence — extract, resolve, compute, detect, and explain SaaS metrics.
Six tools with basic descriptions and partial schemas. All tools have names starting with action verbs (analyze_, import_, compute_, resolve_, detect_, list_) and docstrings. However, parameter descriptions are minimal or absent in most tools. Input schemas are visible but lack depth, most parameters are simple strings with minimal constraint documentation. Output schemas are not formally documented; responses are JSON strings without declared structure. Error handling is absent; no guidance on recovery or retryability. No tool annotations (readOnlyHint, destructiveHint, idempotentHint). Composition is reasonable (each tool has one responsibility), but chaining is implicit, no explicit IDs returned to enable downstream calls.
Full pipeline: import CSV, resolve entities, compute SaaS metrics, detect anomalies.
Compute SaaS metrics (MRR, churn, ARPU, NRR, Quick Ratio) from entity data.
Detect statistical anomalies in SaaS metrics using z-score and rate-of-change methods.
Import SaaS revenue data from a CSV file and return normalized records.
List all available SaaS metric definitions.
Match and deduplicate records across data sources using fuzzy matching. Uses email domain blocking, company name similarity, temporal proximity, and amount matching to resolve which records refer to the same entity.
No output schemas documented. All tools return JSON strings without declared field structure. LLMs cannot plan downstream calls or extract specific fields reliably.
Parameter descriptions are minimal or missing. 'entities_json' and 'records_json' lack format/structure hints. 'period_date' lacks format specification (YYYY-MM-DD is mentioned in docstring but not in schema description).
No error handling or recovery guidance. Tools accept JSON strings but do not validate format or provide actionable error messages if parsing fails.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 54 | <=2025-11-25 | v2 |
No tool annotations. All tools are read-only but lack readOnlyHint. This prevents clients from optimizing caching or retry logic.
Implicit tool chaining. analyze_csv calls import_csv and compute_metrics internally, but the tool interface does not expose intermediate results or IDs for explicit composition.