ir-record

当执行 IR（投资人关系）任务时维护 SQLite 追踪数据库，记录投资人档案、接触历史和项目申报，避免重复，跟踪进展。

Quality

78%

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 ./crews/main/skills/ir-record/SKILL.md

Quality

Content

72%

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

A well-structured, highly actionable DB skill body: executable commands for all operations, clear per-work-block workflows with de-dup guards, and clean one-level-deep bundling into real scripts. Weaker on conciseness (inline schemas overlap the init-db.sh DDL) and workflow clarity (no post-write verification or error-recovery feedback loops, which the rubric expects for database operations).

Suggestions

Trim token redundancy: the inline schema tables repeat the DDL already in init-db.sh — keep the semantic notes and state machines inline but reference init-db.sh for column types, and state the init-db path only once.

Add an explicit verification/feedback loop after writes, e.g. after record-investor.sh run check-investor.sh to confirm the record exists, and document recovery when a status transition is illegal or a write fails.

Surface status-transition validation as a checkpoint in the workflow (e.g. reject moves that violate the new→contacted→…→invested / passed state machines) so database mutations are guarded by explicit validation rather than only pre-write de-dup checks.

Dimension	Reasoning	Score
Conciseness	The body is reference-dense with no concept explanations Claude already knows, but the full column-by-column schema tables duplicate the DDL already in init-db.sh and the init-db path is stated twice (prose + command block), so it could be tightened; it is not a 3 because of that token redundancy, and not a 1 because there is no padding or off-topic explanation.	2 / 3
Actionability	Every script has a copy-paste-ready invocation with all flags, required fields marked ("必填：--name、--type、--firm"), and JSON return formats; all 11 referenced scripts are verified real executable files, matching fully-executable-commands; it is not a 2 because these are complete commands, not pseudocode or abstract guidance.	3 / 3
Workflow Clarity	The 使用规则 section gives clear per-work-block sequences with de-dup pre-checks (check-* before record-*), but there are no explicit post-write verification or error-recovery feedback loops, and the rubric caps database-operation workflows at 2 when those are missing; it is not a 1 because the sequence and pre-check checkpoints are genuinely present, and not a 3 because validation feedback loops are absent.	2 / 3
Progressive Disclosure	The body is a clear overview (schemas, state machines, command catalog, usage rules) pointing to 11 real, one-level-deep bundle scripts signaled by full command paths, with no nested references and easy navigation; scored against the actual bundle structure as the rubric instructs, it is not a 2 because the heavy logic is appropriately split out and references are clearly signaled.	3 / 3
	Total	10 / 12 Passed

Description

85%

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 specific, well-triggered description that clearly states both what the skill does (maintain a SQLite IR tracking DB across investors, contacts, and applications) and when to use it (during IR/投资人关系 tasks). Its only weakness is limited coverage of natural trigger-term variations, which keeps trigger quality at a mid score.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — "维护 SQLite 追踪数据库", "记录投资人档案、接触历史和项目申报", "避免重复", "跟踪进展" — rather than vague language, matching the multiple-specific-actions anchor; it is not a 2 because the action set is comprehensive, not partial.	3 / 3
Completeness	Explicitly answers both: the trigger "当执行 IR（投资人关系）任务时" gives the when, and "维护…数据库，记录…，避免重复，跟踪进展" gives the what; it is not a 2 because the when is an explicit trigger clause, not merely implied.	3 / 3
Trigger Term Quality	Contains the natural trigger "IR（投资人关系）" plus core object terms (投资人/接触/申报), but coverage of common variations is limited — a user might naturally say 融资/跟投/报奖/补贴 without matching, so it does not reach the broad multi-term coverage of 3 and is well above the pure-jargon of 1.	2 / 3
Distinctiveness Conflict Risk	The combination of IR/投资人关系 + SQLite tracking DB + investor profiles/contacts/applications is a clear, narrow niche unlikely to trigger for the wrong skill; it is not a 2 because the trigger and objects are specific enough to avoid meaningful overlap.	3 / 3
	Total	11 / 12 Passed

Validation

100%

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 — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: TeamWiseFlow/xiaobei
Path: crews/main/skills/ir-record/SKILL.md
Commit: bcef823

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.