Atlas / Skills / chekusu / Artifact Naming

Artifact NamingSAFE

skills/chekusu/artifact-naming

wanman is an open-source agent matrix runtime inspired by Japanese one-man trains. It lets human users step back into an observer role while local agent runtimes coordinate autonomous multi-agent workflows, task execution, and artifacts.

Verdict
SAFE
Grade
B
Trust score
89 /100
Version
—
Hosts
—
License
Apache-2.0
Stars
690
01

Overview

wanman is an open-source agent matrix runtime inspired by Japanese one-man trains. It lets human users step back into an observer role while local agent runtimes coordinate autonomous multi-agent workflows, task execution, and artifacts.

Read from source at commit f0c3ec2eb827OBSERVED · 2026-10-13
Advertisement
02

What it tells the agent

The instruction file, verbatim from the audited commit — this is the text the model reads, and the surface the audit's instruction layer examines. Quoted here so you can judge it without cloning anything.

---
name: artifact-naming
description: Naming conventions for artifact kind and path, ensuring data can be aggregated and analyzed
---

# Artifact Naming Conventions

## Important: --path Is a Required Parameter

Every call to `wanman artifact put` must include the `--path` parameter. Artifacts without a path cannot be retrieved or aggregated.

```bash
# Full command format
wanman artifact put --kind <kind> --path "<domain>/<category>/<item>" --source "<source>" --confidence <0-1> '<json>'
```

## Important: Use --file to Store Output File Contents

After writing an MD/HTML or other file, use the `--file` parameter to store the file contents in the artifact. This ensures file contents are version-controlled in the database.

```bash
# After writing a file, include --file in artifact put
wanman artifact put --kind brand_asset --path "brand/identity/handbook" \
  --source "marketing" --confidence 0.9 \
  --file /workspace/agents/output/marketing/brand-design.md \
  '{"name":"Brand Handbook","type":"handbook"}'

# Structured data (no file) does not need --file
wanman artifact put --kind budget_item --path "costs/opex/rent" \
  --source "estimate" --confidence 0.4 \
  '{"item":"rent","amount":350000,"currency":"JPY"}'
```

**Rule:** If your task output includes MD or HTML files, you **must** include `--file` pointing to that file when calling artifact put.

## Kind Standardization

**Only use the following standard kind values — do not invent your own:**

| Domain | Standard Kind | Description |
|--------|--------------|-------------|
| Market research | `competitor` | Competitor information |
| Market research | `market_data` | Market data (size, trends) |
| Market research | `customer_profile` | Customer personas |
| Market research | `location_data` | Location-related data |
| Finance | `budget_item` | Individual cost item (use this uniformly, not monthly_cost/initial_investment) |
| Finance | `revenue_forecast` | Revenue forecast |
| Finance | `financial_summary` | Financial summary (break-even, payback period, etc.) |
| Branding | `brand_asset` | Brand asset (name, story, visuals, tagline — all under brand_asset) |
| Content | `content_plan` | Content planning |
| Technical | `tech_spec` | Technical specification |

## Path Naming Convention (Required)

Format: `{domain}/{category}/{item}`

```
costs/capex/renovation        ← One-time renovation cost
costs/capex/equipment          ← Equipment procurement
costs/opex/rent               ← Monthly rent
costs/opex/labor              ← Monthly labor
costs/opex/materials          ← Raw material costs
market/competitors/cafe-a     ← Competitor A
market/competitors/cafe-b     ← Competitor B
market/demographics/target    ← Target customer segment
market/location/rent-level    ← Rent level
brand/naming/candidate-1      ← Brand name candidate 1
brand/naming/candidate-2      ← Brand name candidate 2
brand/visual/colors           ← Color system
brand/visual/typography       ← Typography direction
brand/story/core              ← Brand story
brand/slogan/main             ← Brand tagline
content/social/instagram      ← Instagram plan
content/social/twitter        ← Twitter plan
revenue/forecast/monthly      ← Monthly revenue forecast
revenue/forecast/breakeven    ← Break-even analysis
finance/summary/cashflow      ← Cash flow summary
tech/website/spec             ← Website technical specification
```

## Correct vs. Incorrect Examples

```bash
# Wrong: kind is too granular, no path
wanman artifact put --kind initial_investment --source "estimate" --confidence 0.4 '{"item":"equipment"}'
wanman artifact put --kind monthly_cost --source "estimate" --confidence 0.4 '{"item":"rent"}'
wanman artifact put --kind brand_name_candidate --source "original" --confidence 0.8 '{"name":"ITTEKI"}'

# Correct: standard kind + path for differentiation
wanman artifact put --kind budget_item --path "costs/capex/equipment" --source "estimate" --confidence 0.4 '{"item":"equipment","amount":500000}'
wanman artifact put --kind budget_item --path "costs/opex/rent" --source "estimate" --confidence 0.4 '{"item":"rent","amount":350000}'
wanman artifact put --kind brand_asset --path "brand/naming/candidate-1" --source "original" --confidence 0.8 '{"name":"ITTEKI","meaning":"one drop"}'
```
03

Trust audit

SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.

LayerWhat it checksResult
L0Provenance & inventoryPASS
L1Static analysis of the codeNA
L2Instruction surface (what it tells the agent)PASS
L3Class-specific surfacePASS
L4Behavioural (sandbox)SKIPPED

What the source does

Filesystem
none-observed
Network
none-observed
Shell
none-observed
Dependencies
pinned
Secrets in source
none-found

Findings (0)

No findings outside the package's declared scope.

Gates applied: no_behavioural_pass.

Audited 2026-10-13 · audit v0.4.1 · source sha f0c3ec2eb827full audit observations/trust-audit/skill/chekusu__artifact-naming.json · Report an issue / request a re-scan
04

Audit history

Every audit this skill has had.

DateSourceVerdictGradeScoreChange
2026-10-13f0c3ec2eb827SAFEB89first audit
05

Questions

What does the Artifact Naming skill do?

wanman is an open-source agent matrix runtime inspired by Japanese one-man trains. It lets human users step back into an observer role while local agent runtimes coordinate autonomous multi-agent workflows, task execution, and artifacts.

Is Artifact Naming safe to install?

The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean skill reads B.

What can Artifact Naming access on my machine?

The audit observed no filesystem, network or shell use at all in its source.

How current is this page?

The grade is for one exact copy of the source (f0c3ec2eb827), read on 2026-10-13. The repository is watched, and a new audit runs when it changes — this is the first audit.

Advertisement