CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

common-subagent-driven-development

Execute an approved implementation plan at the context boundary that preserves required independence and review. Use when implementing plans, executing multi-task PRs, choosing between main-loop and dispatched execution, or transitioning from planning to implementation.

Quality

82%

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

81%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 a disciplined, executable overview: every command is real and verifiable against the bundle scripts, the multi-step workflow has explicit validation gates and failure-specific recovery routes, and detail is properly pushed to one-level-deep references. Remaining gaps are minor: some duplicated maxims, an unnamed script for the status/readme operations, and one cross-bundle reference.

DimensionReasoningScore

Conciseness

The body is extremely dense with no explanations of concepts Claude already knows and no padding. Not 5 because several maxims are repeated nearly verbatim across sections ("prompt text is not enforcement", "never infer permission", "retry count alone never authorizes premium" each appear 2-3 times in Section 1, the Task Loop, Section 4, and Anti-Patterns); not 3 because there is no unnecessary explanation, only minor redundancy.

4 / 5

Actionability

Provides concrete, verified, copy-paste-ready commands with full paths (task_brief.py with required --assignment and --out, review_package.py with $BASE/scope arguments, BASE=$(git rev-parse HEAD)) whose CLI surfaces match the actual scripts. Not 5 because the `status --state ... --format json|markdown` invocation never names its implementing script, `wait()` is referenced without syntax, and the --assignment JSON shape is not shown inline.

4 / 5

Workflow Clarity

A clear 6-step Task Loop (prepare -> dispatch -> yield -> scoped review -> fix loop -> integrate) with explicit validation checkpoints: brief validation that rejects missing contract fields (exit 2), independent review after each task, re-review of the latest cumulative state in the fix loop, terminal-state confirmation before failover, and archive verification before the destructive workspace deletion. This matches the anchor with explicit validation, feedback loops, and error-recovery routing.

5 / 5

Progressive Disclosure

Good structure: the body is an overview with inline deep links to methodology.md anchors that resolve to real headers, and a References section listing all three existing reference files one level deep. Not 5 because sdd_workspace.py is never mentioned anywhere in the body, the `status` command's script is unnamed, and the cross-skill reference ../common-task-complexity-routing/references/routing-table.md is not part of this bundle — minor organization gaps.

4 / 5

Total

17

/

20

Passed

Description

78%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 well-formed description with an explicit, multi-trigger "Use when" clause and a clear statement of what the skill does, in proper third-person imperative voice. Its main weakness is abstraction: it never names the skill's concrete mechanics or the word "subagent", slightly lowering specificity and trigger coverage.

Suggestions

Name one or two concrete operations in the description (e.g., generating task briefs and scoped review packages, dispatching one owner per task) to raise specificity beyond a single abstract action.

Include "subagent" or "delegate/dispatch" as natural trigger terms in the description text so users asking for subagent-based execution match it directly.

DimensionReasoningScore

Specificity

Names the domain ("Execute an approved implementation plan at the context boundary") and roughly two actions, but the operations are abstract — no mention of the concrete mechanics the skill performs (task briefs, dispatch, review packages, receipts). Not 4 because it does not list several specific concrete actions, and not 2 because it goes beyond a minimal generic domain label.

3 / 5

Completeness

Explicitly answers both: what ("Execute an approved implementation plan at the context boundary that preserves required independence and review") and when ("Use when implementing plans, executing multi-task PRs, choosing between main-loop and dispatched execution, or transitioning from planning to implementation") with four concrete trigger phrases. Not 4 because the when-clause is already explicit and specific.

5 / 5

Trigger Term Quality

Contains several natural phrases users would say: "implementing plans", "executing multi-task PRs", "transitioning from planning to implementation". Not 5 because common variations like "subagent", "delegate", or "parallel" never appear in the description text itself; not 3 because coverage clearly exceeds one or two generic keywords.

4 / 5

Distinctiveness Conflict Risk

The niche (plan execution with independence and review gates, "main-loop and dispatched execution") is distinct from general implementation skills. Not 5 because "implementing plans" and "transitioning from planning to implementation" are broad enough to overlap with plan-authoring/planning skills; not 3 because the qualifiers carve out a recognizable specialty.

4 / 5

Total

16

/

20

Passed

Validation

81%

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

Validation — 13 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

metadata_version

'metadata.version' is missing

Warning

metadata_field

'metadata' should map string keys to string values

Warning

relative_links

Relative link issues: 1 suspicious

Warning

Total

13

/

16

Passed

Repository

HoangNguyen0403/agent-skills-standard

Path

.agents/skills/common/common-subagent-driven-development/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.