CtrlK
BlogDocsLog inGet started
Tessl Logo

create-mission-spec

Turn a mission or lab idea into a fully researched GitHub issue spec on microsoft/agent-academy, then post it with the GitHub CLI. Use this skill whenever the user wants to propose new Agent Academy content before writing it — course missions (Recruit, Operative, Commander), standalone Special Ops, or Cowork Collective missions. Trigger whenever the user says "write a spec for a mission", "create an issue for a new lab", "I want to propose a Special Ops", "file a mission proposal", "spec out a lab", "open an issue for a new module", or hands over Microsoft Learn URLs and asks to turn them into an Agent Academy lab. This skill handles: mining source documentation, interviewing for scope, repo reconnaissance and duplicate detection across both docs and existing issues, frontmatter and tag validation, writing-style conformance, drafting the issue body, title conventions, milestone and assignee metadata, and posting plus verifying the issue via `gh`.

72

Quality

90%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Low

Low-risk findings worth noting

SKILL.md
Quality
Evals
Security

Quality

Content

85%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

An excellent single-file operational guide with copy-paste commands, a well-sequenced 8-phase workflow, and validation checkpoints at every risky step (auth gate, duplicate stop, pre-post style check, post-post readback). Its weaknesses are duplicated guidance between the phases and the Rules section, and a ~330-line monolithic body with no reference files despite containing inlineable bulk like the issue-body template and style-constraint tables.

Suggestions

Split inlineable bulk into one-level-deep reference files (e.g. references/issue-body-template.md with the Phase 5 markdown skeleton, and references/style-requirements.md with the per-track style-constraints and screenshot rules), keeping SKILL.md as the phased overview that points to them.

De-duplicate the '📜 Rules' section against the phases: state each rule once (GH_PAGER=cat, projectCards, click/textbox appear in both a phase and Rules) and use Rules only for constraints not covered in any phase.

Trim editorial asides that add length without adding instruction (e.g. 'The difference matters. A spec issue answers what should be built and why. The mission builder answers how it reads. Never write the mission markdown in this skill.') to a single directive line.

DimensionReasoningScore

Conciseness

The body is dense with repo-specific, non-inferable knowledge (fiscal-quarter milestones, GH_PAGER=cat, projectCards deprecation, per-track frontmatter contracts) and avoids explaining concepts Claude already knows. But the '📜 Rules' section restates guidance already given in the phases — GH_PAGER=cat, the projectCards warning, and the click/textbox rule each appear twice — which is trimmable duplication. This fits the 4 anchor (efficient with minor over-explanation) rather than 5 (every token earning its place).

4 / 5

Actionability

Commands are copy-paste ready throughout: the milestones query ('GH_PAGER=cat gh api repos/microsoft/agent-academy/milestones --jq ...'), the full 'gh issue create' invocation with every flag, the verify readback with an exact '--json title,labels,milestone,assignees,url' field list, plus exact title formats and a complete issue-body template. This matches the 5 anchor — fully executable guidance covering the common cases.

5 / 5

Workflow Clarity

Eight explicitly numbered phases with checkpoints at every fragile point: a prerequisites gate ('Do not proceed until both checks pass'), a stop-and-ask on duplicate matches, a style check run before posting ('Fix anything the check surfaces before posting. Do not post and then clean up.'), user confirmation before the irreversible public action, and a Phase 8 readback that verifies all four metadata fields and H1/title consistency. This matches the 5 anchor — clear sequence, explicit validation, and feedback loops for error recovery.

5 / 5

Progressive Disclosure

There are no bundle files at all (no references/, scripts/, or assets/), and the ~330-line body inlines substantial material that belongs in one-level-deep reference files: the complete issue-body template, the per-track style-constraints table, and the screenshot rules. Section headers and tables do give it real structure, and the files it does point to (WRITING_STYLE.md, CONTRIBUTING-*.md) are runtime targets in another repo rather than bundle references. That places it at the 3 anchor ('some structure; content that should be separate is inline') rather than 4, which would require most bulk content to be appropriately split out.

3 / 5

Total

17

/

20

Passed

Description

96%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A dense, well-engineered description: explicit what and when, third-person voice, exhaustive concrete action list, and verbatim natural-language triggers. The only weakness is that a couple of the quoted trigger phrases are generic enough to collide with generic issue-filing requests outside this repo.

DimensionReasoningScore

Specificity

The description lists ten concrete actions ('mining source documentation, interviewing for scope, repo reconnaissance and duplicate detection across both docs and existing issues, frontmatter and tag validation, writing-style conformance, drafting the issue body, title conventions, milestone and assignee metadata, and posting plus verifying the issue via `gh`'), covering the full workflow end to end. It matches the 5 anchor's comprehensive multi-action coverage; there are no unnamed capability gaps in the workflow it claims.

5 / 5

Completeness

Both halves are explicit: 'what' is stated in the opening sentence and the 'This skill handles:' list, and 'when' is stated twice ('Use this skill whenever the user wants to propose new Agent Academy content before writing it' and 'Trigger whenever the user says...'). This matches the 5 anchor (explicit what AND when with concrete trigger phrases) rather than the 4 anchor where the 'when' is only loosely specified.

5 / 5

Trigger Term Quality

Seven verbatim natural trigger phrasings ('write a spec for a mission', 'create an issue for a new lab', 'I want to propose a Special Ops', 'file a mission proposal', 'spec out a lab', 'open an issue for a new module') plus a URL-handover case give comprehensive coverage with synonyms across mission/lab/module/proposal/issue vocabulary. This matches the 5 anchor's coverage including natural variations; a 4 would require missing common phrasings, and none are evident.

5 / 5

Distinctiveness Conflict Risk

The niche is clear — GitHub issue specs for Agent Academy missions on microsoft/agent-academy — with track-specific vocabulary (Recruit, Operative, Commander, Special Ops, Cowork Collective) that is distinct. However, generic triggers like 'open an issue for a new module' and 'create an issue for a new lab' could fire for a user wanting any GitHub issue filed on any repo, giving minor overlap risk with generic issue-creation skills — the 4 anchor rather than the 5 anchor's 'minimal conflict risk'.

4 / 5

Total

19

/

20

Passed

Validation

93%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

15

/

16

Passed

Repository
microsoft/agent-academy
Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.