CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

context-mode-ops

Manage context-mode GitHub issues, PRs, releases, and marketing with parallel subagent army. Orchestrates 10-20 dynamic agents per task. Use when triaging issues, reviewing PRs, releasing versions, writing LinkedIn posts, announcing releases, fixing bugs, merging contributions, validating ENV vars, testing adapters, or syncing branches.

Quality

67%

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

Fix and improve this skill with Tessl

tessl review fix ./.claude/skills/context-mode-ops/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

31%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 operational core (blocking gates, workflow detection table, gh CLI rules, validation checklist) is genuinely good, but it is buried under a triple-stated owner directive that consumes most of the file, and every detailed workflow lives in referenced files that are not included in the bundle. The skill also carries non-operational content (MRR pressure, Codex competition taunts) and an unresolved contradiction between mandatory delegation and mandatory first-hand verification.

Suggestions

Collapse the triple-stated directive (owner monologue + 12 decoded principles + 12 MUST rules) into one concise rules section, and move any verbatim directive text you must preserve into a single reference file — the current ~250-line preamble restates the same ~40 lines of policy three times.

Include the eight referenced files (validation.md, tdd.md, triage-issue.md, review-pr.md, release.md, marketing.md, agent-teams.md, communication.md) in the bundle, or inline minimal essentials for each workflow — currently every workflow link in the detection table is dead.

Remove non-operational framing (MRR/sales pressure, the Codex-competition taunt) and resolve the contradiction between 'you are FORBIDDEN from reading source code' and the file:line citation requirements by stating explicitly that subagents perform the reads and the EM verifies their citations.

DimensionReasoningScore

Conciseness

Roughly 250 of ~440 lines are an "OWNER OPERATING DIRECTIVE" preamble that states the same rules three times — a verbatim owner monologue, then 12 "Decoded operating principles", then 12 "MUST rules" covering identical ground (e.g. principle #5 and MUST-8 both restate the TDD law verbatim in spirit). This matches the "severely verbose; heavily padded" anchor; the lean operational sections at the end are not enough to lift it to 2, since the padding is duplication rather than trimmable over-explanation.

1 / 5

Actionability

There is real executable guidance (gh commands like "gh pr merge --squash", validation commands like "npx vitest run tests/adapters/" and "npm run typecheck", the clone recovery command "git clone --depth 1 <url> refs/platforms/<name>"), but key workflow steps remain abstract ("Ping-pong — Route Architect reviews ↔ Staff Engineer fixes", "wear your PO hat", "Build agent roster... context-driven") and the concrete per-workflow instructions are delegated to files not present in the bundle. This lands between the pseudocode-level and mostly-executable anchors.

3 / 5

Workflow Clarity

The top-level EM flow (Analyze → Recruit → Dispatch → Ping-pong → Ship) is sequenced and blocking gates with a validation checklist exist, but the actual triage/review/release/marketing workflows have no in-file sequence — they are only a pointer table to triage-issue.md, review-pr.md, release.md, and marketing.md, none of which ship in this bundle. Checkpoints are also undercut by an internal contradiction: the delegation gate forbids the main agent from "reading source code" while MUST-6/claim-verification demand first-hand file:line evidence.

3 / 5

Progressive Disclosure

All eight referenced files (validation.md, tdd.md, triage-issue.md, review-pr.md, release.md, marketing.md, agent-teams.md, communication.md) are missing from the bundle — the bundle contains no reference files at all, so every deep link is dead. Simultaneously, large content that clearly belongs in separate files is inlined: the 250-line directive, the 17-adapter enumeration, and the refs/platforms upstream-repository table. This matches the anchor for inlined content that belongs in separate files plus broken navigation; it is not 3 because the missing files make the otherwise well-signaled reference structure non-functional.

2 / 5

Total

9

/

20

Passed

Description

88%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 strong description: third-person voice, concrete multi-action scope, and an explicit "Use when..." clause with natural trigger phrases that mirror the body's workflow detection table. The only weaknesses are a slightly buzzwordy "parallel subagent army" framing and a few missing trigger synonyms that the body already uses.

DimensionReasoningScore

Specificity

The description lists many concrete actions — "triaging issues, reviewing PRs, releasing versions, writing LinkedIn posts, announcing releases, fixing bugs, merging contributions, validating ENV vars, testing adapters, or syncing branches" — comprehensively covering the skill's ops surface. It clearly matches the anchor for multiple specific concrete actions with comprehensive coverage; it is not score 4 because there are no meaningful coverage gaps in the action list.

5 / 5

Completeness

It explicitly answers both questions: "what" ("Manage context-mode GitHub issues, PRs, releases, and marketing with parallel subagent army. Orchestrates 10-20 dynamic agents per task") and "when" ("Use when triaging issues... or syncing branches") with concrete trigger phrases. This is a direct match to the top anchor; there is no missing or weakly implied half.

5 / 5

Trigger Term Quality

Trigger phrases are natural verb phrases users would actually say ("reviewing PRs", "fixing bugs", "releasing versions", "writing LinkedIn posts", "validating ENV vars", "syncing branches") and align with the body's workflow table. It falls just below the comprehensive anchor because synonyms and variations present in the body (e.g. "version bump", "publish", "merge PR", "triage issue #N") are absent from the description itself.

4 / 5

Distinctiveness Conflict Risk

The "context-mode" project scope, LinkedIn/marketing angle, and adapter/ENV-var testing terms carve a clear niche with distinct triggers. It is not fully a 5 because generic phrases like "fixing bugs", "reviewing PRs", and "merging contributions" overlap with general code-review and git workflow skills, creating minor conflict risk with closely related skills.

4 / 5

Total

18

/

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

relative_links

Relative link issues: 15 missing

Warning

Total

15

/

16

Passed

Repository

mksglu/context-mode

Path

.claude/skills/context-mode-ops/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.