CtrlK
BlogDocsLog inGet started
Tessl Logo

getlarge/legreffier-explore

Systematic diary exploration: discover tags, entry distribution, coverage gaps, agent mistakes, and compile recipes

86

1.06x
Quality

90%

Does it follow best practices?

Impact

81%

1.06x

Average score across 5 eval scenarios

SecuritybySnyk

Advisory

Suggest reviewing before use

Overview
Quality
Evals
Security
Files

criteria.jsonevals/scenario-1/

{
  "context": "Tests whether the agent implements the full exploration output format, recovery protocol, and relation opportunity detection: report structure with all required sections, diary entry metadata (reflection type, exploration+diary-health tags, importance 6), context compression recovery, and cross-type relation candidates.",
  "type": "weighted_checklist",
  "checklist": [
    {
      "name": "All report sections",
      "description": "Report template includes all required sections: Inventory, Tag Namespaces, Agent Mistakes, Commit Patterns, Coverage Gaps, Noise Sources, Recommended Compile Recipes",
      "max_score": 12
    },
    {
      "name": "Entry type reflection",
      "description": "Report or code specifies that the exploration output should be stored as entry_type: reflection",
      "max_score": 8
    },
    {
      "name": "Exploration tags",
      "description": "Report metadata uses tags exploration and diary-health",
      "max_score": 8
    },
    {
      "name": "Importance 6",
      "description": "Report metadata sets importance to 6",
      "max_score": 6
    },
    {
      "name": "Recovery detection",
      "description": "Recovery protocol searches for existing exploration entry using the exploration tag before starting a new exploration",
      "max_score": 12
    },
    {
      "name": "Phase resumption",
      "description": "Recovery protocol reads the existing exploration entry, determines which phases are complete, and resumes from the next incomplete phase",
      "max_score": 10
    },
    {
      "name": "Incident-antipattern relations",
      "description": "findRelationOpportunities identifies incidents that prove scan entry anti-patterns as relation candidates",
      "max_score": 8
    },
    {
      "name": "Decision-commit relations",
      "description": "findRelationOpportunities identifies decisions referenced by procedural commits as relation candidates",
      "max_score": 8
    },
    {
      "name": "Repeated incident relations",
      "description": "findRelationOpportunities identifies repeated incidents (same bug pattern across branches) as relation candidates",
      "max_score": 8
    },
    {
      "name": "Header fields",
      "description": "Report header includes date, total entries analyzed, and temporal range",
      "max_score": 8
    },
    {
      "name": "Severity grouping in report",
      "description": "Agent Mistakes section groups incidents by severity (Critical/High vs Medium/Low)",
      "max_score": 6
    },
    {
      "name": "Ordered phases",
      "description": "Report or code documents that exploration phases must run in order (1-5), each building on previous findings",
      "max_score": 6
    }
  ]
}

evals

scenario-1

criteria.json

task.md

tile.json