agentsclimarketplace

Publish public

Skill pantheon-org/tekhne/skills/agentic-harness/tessl/publish-public

Agents Skills

Install
npx -y skills add pantheon-org/tekhne --skill publish-public

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 9 stars9 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Ensure Tessl plugins meet all requirements for public registry publication with comprehensive validation. Use when publishing skills to public registry, validating .tessl-plugin/plugin.json configuration, creating evaluation scenarios, checking quality thresholds (A-grade >=108/120), or preparing plugins for release. Validates eval scenario coverage, plugin.json fields (name, version, private, description, skills), agent-agnostic compliance, and publication readiness.

SKILL.md

4.3 KB, as published. Nobody here has run it

Tessl Public Publication Skill

Ensure Tessl plugins meet all requirements for public registry publication, including evaluation scenario coverage, quality thresholds, and proper plugin configuration.

Principles

Public plugins represent a quality commitment to every agent that installs them. Apply this skill when you need to validate, prepare, or execute a public registry publication.

When to use: Publishing a plugin to the public Tessl registry, verifying publication readiness, or enforcing quality gates before release.

When not to: Private workspace publishing, internal-only plugins, or pre-alpha prototypes where A-grade quality is not yet expected.

The three non-negotiable gates are:

  1. Quality audit passes (>=108/120 A-grade via skill-quality-auditor)
  2. Evaluation scenarios exist (minimum 5 scenarios with measurable success criteria)
  3. plugin.json is correctly configured (private: false, valid fields)

Workflow

1. Pre-Publication Quality Audit

sh skills/agentic-harness/skill-quality-auditor/scripts/evaluate.sh <domain>/<skill-name> --json --store
cat .context/audits/<domain>/<skill-name>/latest/analysis.md

If below A-grade, review the remediation plan and re-audit after improvements.

2. Create Evaluation Scenarios

mkdir -p skills/<domain>/<skill-name>/evals/scenario-01

Each scenario needs: task.md, criteria.json (checklist summing to 100), capability.txt. Target 5-8 scenarios covering the full trigger surface.

3. Configure .tessl-plugin/plugin.json

{
  "name": "workspace/skill-name",
  "version": "1.0.0",
  "private": false,
  "description": "Descriptive text with use cases. Keywords: term1, term2",
  "skills": ["SKILL.md"]
}

All 5 fields are required. private: false must be boolean, not string. skills is an array of path strings.

See references/plugin-json-schema.md for field validation rules and examples.

4. Run Tessl Review

tessl review run skills/<domain>/<skill-name>
tessl review run skills/<domain>/<skill-name> --optimize

Target >=90% from Tessl (independent of skill-quality-auditor).

5. Validate Agent-Agnostic Compliance

Prohibited: harness-specific tool references, IDE-specific commands, platform-specific paths. Required: universal tools only (Read, Write, Edit, Bash, Grep, Glob).

6. Publish

sh scripts/check-publication-readiness.sh skills/<domain>/<skill-name>
tessl plugin publish --workspace pantheon-ai skills/<domain>/<skill-name> --bump patch
tessl search <skill-name>

Anti-Patterns

  • NEVER skip evaluation scenarios -- WHY: public registry requires proof of effectiveness via measurable scenarios
  • NEVER publish below A-grade (108/120) -- WHY: sub-threshold plugins erode registry quality and may be flagged
  • NEVER set private: true -- WHY: plugin.json defaults to private; must be explicitly set to false
  • NEVER skip --optimize when below 90% -- WHY: optimization routinely lifts scores from 85% to 99%
  • NEVER use harness-specific tool calls -- WHY: public plugins must work across all agent platforms

See references/anti-patterns.md for detailed examples with remediation steps.

Gotchas

  • Eval scenarios are NOT optional: Public plugins require evaluation scenarios
  • Quality audit != Tessl review: Both are independent checks; both must pass
  • Version bumping required: Republishing requires a version increment via --bump
  • No rollback: Published versions are immutable; only newer versions supersede

References

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.