CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

claude-code-bridge

Integrate with Claude Code CLI for delegating complex coding tasks. Use when you need to spawn a separate Claude Code instance to work on files, run tests, or perform multi-step coding operations in a specific directory. Enables "Claude calling Claude" patterns for parallel work or specialized contexts.

Quality

72%

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

Fix and improve this skill with Tessl

tessl review fix ./planning/done/claude-code-integration/SKILL.md

The canonical home for this skill is claude-code-bridge in Raudbjorn/TTTRPS

SKILL.md
Quality
Evals
Security

Quality

Content

53%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 well-organized and reasonably lean, with a genuinely executable CLI section as its backbone. Its weaknesses are duplicated example workflows, non-executable pseudo-notation for the recommended MCP path, an unverifiable Rust library example, and validation guidance that stays advisory rather than being embedded as workflow checkpoints.

Suggestions

Show MCP usage with concrete, executable syntax (actual tool invocations or a session transcript) instead of '[claude_prompt]' pseudo-blocks, and remove or verify the Rust example — if the claude_code_rs crate is not a real published dependency, replace it with a real integration path.

Trim the Example Workflows section to patterns that add value beyond Usage Patterns (e.g., keep only the multi-turn followup workflow) and cut narrative filler lines like 'Let me have Claude Code generate tests:'.

Embed validation checkpoints directly in the workflows (e.g., 'after claude_prompt returns, inspect the result before continuing the conversation') instead of listing them as advisory bullets in Best Practices.

DimensionReasoningScore

Conciseness

The body is mostly efficient — terse command blocks, a compact best-practices list, and no explanations of concepts Claude already knows — but the Example Workflows section (~30 lines) largely duplicates the Usage Patterns section, and narrative filler like 'Let me have Claude Code generate tests:' adds tokens without information. This matches the level-3 anchor (mostly efficient but could be tightened) better than level 4, where over-explanation would be only minor.

3 / 5

Actionability

The CLI section is fully executable ('claude -p "..." --output-format json', '--continue', '--resume <session-id>'), but the recommended MCP path is shown only in non-executable pseudo-notation ('[claude_prompt]' blocks), and the Rust snippet references 'claude_code_rs', a crate that is not a verifiable real dependency, reading closer to pseudocode. This sits between the level-3 anchor (concrete but incomplete, pseudocode instead of executable code) and level 4, settling at 3 because two of the three usage paths are not copy-paste ready.

3 / 5

Workflow Clarity

The usage patterns and the 'Refactoring with Followup' example establish a rough sequence including a review-then-continue feedback loop, and 'Check Results: Verify Claude Code's output before proceeding' names a checkpoint. However, validation exists only as an advisory bullet in Best Practices rather than explicit checkpoints embedded in the workflows, matching the level-3 anchor (sequence present but checkpoints implicit) rather than level 4.

3 / 5

Progressive Disclosure

There are no bundle files, and the ~135-line body is organized into clear, well-labeled sections (When to Use, Usage Patterns, Configuration, Best Practices, Example Workflows, Error Handling) with no nested or buried references. This fits the level-4 anchor (good structure, most content appropriately placed, minor organization gaps): the MCP configuration variants and example workflows could arguably be split into reference files, which is the minor gap keeping it from 5.

4 / 5

Total

13

/

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.

The description is strong: it clearly states both what the skill does and when to use it, with concrete trigger phrases and mostly natural terminology. The main deductions are the second-person voice in the 'when' clause and the broad 'delegating complex coding tasks' framing, which slightly blurs distinctiveness and keyword coverage.

DimensionReasoningScore

Specificity

The description lists several concrete actions ('work on files, run tests, or perform multi-step coding operations in a specific directory'), which matches the level-4 anchor, but the second-person phrasing 'Use when you need to spawn' triggers the rubric's third-person voice penalty of -1, landing it at 3. It is not a 2 because the domain and actions are specific rather than generic, and not a 4 because the voice penalty explicitly applies.

3 / 5

Completeness

It explicitly answers both 'what' ('Integrate with Claude Code CLI for delegating complex coding tasks') and 'when' ('Use when you need to spawn a separate Claude Code instance to work on files, run tests, or perform multi-step coding operations in a specific directory') with concrete trigger phrases. This clearly matches the level-5 anchor rather than the level-4 anchor, whose 'when' clause is less specific.

5 / 5

Trigger Term Quality

Natural trigger terms like 'delegating complex coding tasks', 'spawn a separate Claude Code instance', and 'parallel work' match what a user would say, giving good keyword coverage. It falls short of 5 because common variations such as 'subagent', 'background task', or 'run another Claude session' are missing, and 'Claude calling Claude' is jargon rather than a user phrase.

4 / 5

Distinctiveness Conflict Risk

The niche of spawning a separate Claude Code CLI instance is distinct, but the broad phrase 'delegating complex coding tasks' creates minor overlap risk with general agent/subagent or delegation skills, matching the level-4 anchor. It is not 5 because the opening framing alone could plausibly match other delegation-oriented skills, and not 3 because the specific trigger language keeps it well differentiated.

4 / 5

Total

16

/

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

Raudbjorn/TTTTRPS

Path

planning/done/claude-code-integration/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.