Install
$ agentstack add skill-45ck-skill-harness-spec-writer ✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.
Security review
✓ PassedNo 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.
About
Write a spec document for the feature or component described. Place it in docs/ with a .md extension.
Spec sources: docs/ is the primary directory. specs/ is a legacy fallback.
Required frontmatter fields
---
id: # unique ID, e.g. AUTH-001
title: "Human-readable title"
state: draft | proposed | in_progress | accepted | done | deprecated
kind: functional | non_functional | architecture | interface | constraint
required_evidence:
implementation: E0 # minimum evidence strength needed
# add more dimensions as needed: verification, models, apis
# optional:
depends_on: [OTHER-001] # IDs this spec depends on
conflicts_with: [OTHER-002] # IDs this spec conflicts with
owner: "Team or person"
tags: [auth, security]
---
Evidence strength reference
| Level | Name | Source | |-------|------|--------| | E0 | Declarative | @spec/@implements JSDoc annotation | | E1 | Structural | Beads closed issue; file/symbol reference | | E2 | Indexed | Stored in specgraph DB; linked artifact | | E3 | Automated | Passing test suite with @spec annotation | | E4 | Runtime | CI artifact, coverage report, live probe |
State guidance
| State | Policy | Notes | |---|---|---| | draft | Always passes | No evidence required — early exploration | | proposed | Advisory | Warns if no implementation found | | in_progress | E1 implementation required | Verification is advisory | | accepted | E2 impl + E2 verification + E1 models | Full evidence required | | done | E3 impl + E3 verification + E2 models | High-confidence evidence | | deprecated | Always passes | No longer enforced |
Evidence dimensions
The required_evidence block accepts any combination of:
| Dimension | What it tracks | |---|---| | implementation | Code that implements the spec | | verification | Tests that verify the spec | | models | Data models referenced by the spec | | apis | API endpoints implementing the spec |
Inline waiver format (frontmatter)
waivers:
- kind: missing-verification
target: "*"
owner: eng-lead
reason: "No automated tests yet — deferred to milestone 3"
expires: "2026-12-31"
Do not add waivers to the frontmatter directly for complex cases — use specgraph waivers or standalone docs/waivers/-.md files instead.
Rules
- Choose the smallest evidence requirement that still provides meaningful signal.
- Keep
idstable — downstream annotations and Beads issues reference it. - Write section body in imperative: "The system SHALL…", "The API MUST…".
- Add
depends_onwhen this spec only makes sense after another is satisfied. - Add
conflicts_withwhen this spec is mutually exclusive with another. - Place spec files in
docs/— this is the primary scanned directory.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: 45ck
- Source: 45ck/skill-harness
- License: MIT
- Homepage: https://github.com/45ck/skill-harness/releases/tag/v0.1.0
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.