CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

apm-2-initiate-manager

Initiate an APM Manager. Use only when the User manually invokes this skill.

Quality

83%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

100%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.

The body is an exemplary instruction-only skill: lean, assumes competence, and concrete at every step with exact file paths, skill names, and section pointers. Multi-step workflows carry explicit validation checkpoints and feedback loops, and detail is appropriately pushed to clearly signaled one-level-deep guides.

DimensionReasoningScore

Conciseness

The body is dense procedural instruction with zero concept explanations — git, worktrees, buses and compaction are all assumed ("Watch for signs that a Worker is losing context: it says its context was compacted, repeats finished work, ignores the Rules"). Every line directs behavior; no padded sections exist, matching the lean anchor-5 profile.

5 / 5

Actionability

For an instruction-only skill the guidance is fully concrete: exact paths (".apm/bus/manager/handoff.md", "AGENTS.md"), exact skill names to invoke ("apm-5-check-reports", "apm-8-summarize-session"), precise section pointers ("{GUIDE_PATH:task-assignment} §1"), and explicit decision branches ("If it has content, continue with §1.2; otherwise this is the first Manager, so continue with §1.1"). Per the rubric's instruction-skill note, the absence of code is not penalized.

5 / 5

Workflow Clarity

Sequences are numbered with explicit validation checkpoints and feedback loops: "Wait for approval, apply corrections and ask again", "Check what they say about branches, merges, Worker buses and pending reviews against the Tracker, the bus files and, with version control, git. Where they differ, trust the current files and git", and "Wait for confirmation that every active Worker has stopped writing" before recording state — the anchor-5 pattern including error-recovery guidance.

5 / 5

Progressive Disclosure

The body stays a role overview and delegates the mechanics one level deep through clearly signaled references ("{GUIDE_PATH:task-review}", "{SKILL_PATH:apm-communication}", "apm-6-handoff-manager"), with clean section structure (Initiation, Coordination Cycle, Session Completion, Subagents). No bundle files exist to inline instead, and no content that belongs elsewhere is embedded — matching the anchor-5 structure.

5 / 5

Total

20

/

20

Passed

Description

57%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.

The description is short, in third person, and explicitly gates usage, which suits a manual-invocation skill, but it says almost nothing about what the Manager actually does. Specificity and trigger coverage are the weak points; both 'what' and 'when' are present, keeping completeness solid.

Suggestions

Add 2-3 concrete actions to raise specificity, e.g., "Initiate an APM Manager: read the Tracker, Plan and Spec, dispatch Tasks to Workers, review Task Reports and keep the Tracker and Memory current."

Include natural trigger phrasing even for manual invocation, e.g., "Use when the user asks to start, take over, or re-initiate the APM Manager (e.g., 'start the manager', 'continue as Manager')."

Sharpen distinctiveness by naming the boundary with sibling skills, e.g., "Use apm-6-handoff-manager to hand off a running Manager, not this skill."

DimensionReasoningScore

Specificity

"Initiate an APM Manager" names the domain but relies on a single generic action — structurally the same as the anchor-2 example "Processes PDF files". It does not reach anchor 3 because no concrete actions of the Manager (dispatching Tasks, reviewing reports, maintaining the Tracker) are stated.

2 / 5

Completeness

It has both parts: "Initiate an APM Manager" (what) and "Use only when the User manually invokes this skill" (an explicit when, so not capped at 3). It falls short of anchor 5 because the 'what' is a single thin action and the 'when' offers no concrete trigger phrases beyond the manual-invocation restriction.

4 / 5

Trigger Term Quality

"APM Manager" and "manually invokes" are relevant, non-generic keywords a user of this system would naturally say, but no variations or synonyms (e.g., "start", "coordinator", skill-name mention) appear. This matches anchor 3 — some relevant keywords, missing common variations — rather than anchor 4's good coverage.

3 / 5

Distinctiveness Conflict Risk

"APM Manager" carves a clear niche and the "Use only when the User manually invokes this skill" gating limits misfires, but the APM family contains closely related Manager skills (e.g., apm-6-handoff-manager), leaving minor overlap risk — anchor 4 rather than 5's minimal-conflict profile.

4 / 5

Total

13

/

20

Passed

Validation

100%

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

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository

sdi2200262/agentic-project-management

Path

templates/skills/apm-2-initiate-manager/SKILL.md

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.