data-routing

The single ROUTER for every data need. Load this skill BEFORE any backtest, data-fetch, or research task to pick the best available source/tool, honour auth (env) requirements, and avoid ban-risk providers.

Quality

77%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./agent/src/skills/data-routing/SKILL.md

Quality

Content

75%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A well-built routing reference: dense lookup tables, explicit env-key gating, fallback and error-envelope rules, and a symbol-format spec make it genuinely actionable. The main gaps are mild token overhead in the explanatory notes and a large inlined source table that would serve better as a separate reference file.

Suggestions

Move the detailed Source Overview and Symbol Format tables into a references/ file (e.g. references/sources.md), keeping SKILL.md to the routing rules, decision tree, and a compact source-priority summary.

Trim the implementation-detail notes under the capability table and the closing "This discipline is what separates analysis from aggregation" line; keep only behavior-relevant facts (error envelopes, throttling).

Add "longbridge" to the Source Overview (it appears in the HK priority chain with a key-gate note but is absent from the source table), and show one concrete tool invocation example so the guidance is copy-paste ready.

Dimension	Reasoning	Score
Conciseness	Dense, high-signal tables that assume competence, with only minor trims available: the "Every name here is a registered OHLCV/backtest source" sentence repeats the intro, and the implementation-detail notes plus the closing flourish "This discipline is what separates analysis from aggregation" add tokens without routing value.	4 / 5
Actionability	Gives concrete tool names, exact env keys, worked symbol examples (TD.TO, BTC-USDT, CU2406.SHFE), ordered priority chains, and the cross_validate argument shape — actionable but not fully copy-paste ready since tool invocation syntax is never shown and "longbridge" appears in a priority chain without being in the source table.	4 / 5
Workflow Clarity	The research flow is numbered with an explicit checkpoint ("report the missing key rather than failing silently") and fallback loops exist (throttle/timeout fallback, per-symbol envelope errors that never abort a batch). Not a 5 because the backtest branch is a single unstructured line and the cross-check adoption step is partially implicit.	4 / 5
Progressive Disclosure	Well-organized sections with a clear overview and no nested references, but the 24-row source table and symbol-format reference are dense lookup material that could live in a one-level-deep reference file to keep the SKILL.md overview leaner.	4 / 5
	Total	16 / 20 Passed

Description

71%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A solid router description with an explicit "Load BEFORE..." trigger and three concrete routing actions, weakened by an abstract "single ROUTER for every data need" framing that both understates the concrete domain and overstates the scope. More specific trigger vocabulary (market data, OHLCV, price data) would sharpen it.

Suggestions

Replace the abstract "The single ROUTER for every data need" framing with a concrete statement of what is routed, e.g. "Routes market-data needs (OHLCV price bars, fund flow, financial statements, news) to the right data source and tool for each market."

Broaden trigger vocabulary with natural user phrasings such as "market data", "OHLCV", "price data", or "fetch stock data" alongside backtest/data-fetch/research.

Narrow the universal "every data need" claim to the actual coverage (backtest and research data sources) to reduce overlap with other data-handling skills.

Dimension	Reasoning	Score
Specificity	Names three concrete actions ("pick the best available source/tool, honour auth (env) requirements, and avoid ban-risk providers") but never states what is actually routed — markets, data types, or tool names — so coverage has minor gaps rather than being comprehensive.	4 / 5
Completeness	Both sides are present: the what is given as routing actions and the when is an explicit trigger clause ("Load this skill BEFORE any backtest, data-fetch, or research task"). Not a 5 because the "what" leans on the abstract phrase "The single ROUTER for every data need" rather than concrete capability statements.	4 / 5
Trigger Term Quality	Includes natural terms users would say ("backtest", "data-fetch", "research task") with good coverage, but misses common variations like "market data", "OHLCV", "price data", or "stock data".	4 / 5
Distinctiveness Conflict Risk	The claim "The single ROUTER for every data need" is maximally broad and overlaps with any data-related skill; the backtest/auth/ban-risk framing is domain-specific, keeping it above anchor 2, but the universal-scope phrasing invites wrong-skill triggering.	3 / 5
	Total	15 / 20 Passed

Validation

93%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: HKUDS/Vibe-Trading
Path: agent/src/skills/data-routing/SKILL.md
Commit: 251b094

Reviewed: 3 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

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.