AgentStack
Sign in Publish
MCP servers Skills Toolkits Why AgentStack Sell Docs
Sign in
SKILL verified MIT Self-run

Workspace Memory Benchmark Authoring

skill-fightzy-simple-skills-workspace-memory-benchmark-authoring · by fightZy

Use when the user wants to add or expand workspace-memory benchmark fixture coverage, especially for complex retrieval cases that can be drafted safely by subagents under the benchmark fixture directory.

↗ Repository
— No reviews yet
0 installs
2 views
0.0% view→install

Install

$ agentstack add skill-fightzy-simple-skills-workspace-memory-benchmark-authoring

✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.

Security review

✓ Passed

No issues found. Passed automated security review. · v0.1.0 How review works →

  • ✓ Prompt-injection patterns
  • ✓ Secret / credential exfiltration
  • ✓ Dangerous shell & filesystem operations
  • ✓ Untrusted network calls
  • ✓ Known-malicious package signatures

What it can access

  • ✓ Network access No
  • ✓ Filesystem access No
  • ✓ Shell / process execution No
  • ✓ Environment & secrets No
  • ✓ Dynamic code execution No

From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.

Are you the author of Workspace Memory Benchmark Authoring? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

Workspace Memory Benchmark Authoring

Use this companion skill when the job is to add or refine self-built benchmark fixtures for workspace-memory-skill.

Scope

This skill is for repository benchmark authoring, not for runtime memory operations.

Primary paths:

  • tests/workspace-memory-skill/benchmark_fixtures/
  • tests/workspace-memory-skill/test_workspace_memory_benchmark.py
  • scripts/benchmarks/workspace_memory/

Workflow

  1. Inspect existing fixture coverage before adding new cases.
  2. Identify missing buckets or edge conditions.
  3. Keep each fixture focused on one retrieval behavior.
  4. If fixture drafting is independent, use a subagent restricted to benchmark_fixtures/ only.
  5. Review fixture realism locally before accepting it.
  6. Run benchmark verification after fixture changes.

Fixture Rules

Each fixture should define:

  • case_id
  • query
  • memory_files
  • expected_candidate_ids
  • bucket

Add these when they help:

  • expected_top_id
  • expected_layers
  • forbidden_candidate_ids
  • gold_answer
  • qa_enabled

Keep fixture files realistic:

  • use frontmatter-rich Markdown
  • use stable ids and tags
  • keep memory trees small but believable
  • avoid testing multiple ranking ideas in one case

Subagent Boundary

If you delegate fixture drafting:

  • the subagent owns only tests/workspace-memory-skill/benchmark_fixtures/
  • do not let the subagent edit runner, scorer, or runtime scripts
  • review every drafted case locally before verification

Verification

Run:

  • python -m pytest tests/workspace-memory-skill/test_workspace_memory_benchmark.py -q
  • python -m scripts.benchmarks.workspace_memory.runner tests/workspace-memory-skill/benchmark_fixtures

If a fixture exposes a real retrieval bug, report that separately instead of silently weakening the case.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet — be the first.

Write a review

Versions

  • v0.1.0 Imported from the upstream source.

More in Search