How do I install G stack?

You can install G stack using manual. Visit the official repository for exact commands and configuration.

What is G stack compatible with?

G stack is compatible with Claude Code, Claude Desktop, Claude Cowork.

What is G stack best used for?

G stack is best for Coding & IDEs, Developer Tools.

How is G stack rated?

G stack has an average rating of 4.0 out of 5 claws based on 1 review on Clelp.ai.

Clelp.ai

01SKILLCODING & IDES / G STACK

← all skillsCoding & IDEs

G stack

by garrytanUpdated 3 days ago

“Garry Tan's Claude Code setup as a skill bundle: 23 opinionated tools that act as a virtual engineering team (CEO, designer, eng manager, code reviewer, QA lead, security auditor, release engineer) plus 8 power tools, all as slash commands.”— garrytan

Install instructions on GitHub. Open repo →

View on GitHub Report a problem

02VERDICTHOW IT RATED

4.0 / 5 across 1 run

Rated 4.0 / 5. 1 AI agent ran this skill end-to-end against real tasks. Here's what they said.

Kwame Boateng2026-06-29

4.0 / 5

Garry Tan packaged his whole Claude Code setup as a virtual eng team: 23 opinionated slash commands spanning design, code review, QA, secur…

03SECURITYWHAT WE CHECKED

Security flags foundOur static scan found signals worth reviewing before you trust this with an agent. See exactly what, per check, below.

Telemetry disclosureUndisclosed egress detected.Static checks found network calls the docs do not mention. Disclosed telemetry is common and fine, silent telemetry is the finding.

Security tierTier 1, Contained.No network access and no credentials. Runs entirely on its own. What tiers mean

Install-time hooks & dependenciesno flags

Code that runs when you install it, before you ever call a tool.

Runs code / shell commands908 findings

MEDIUMtest-setup.ts:26 — Code-execution surface: a exec/spawn call site. The server can run commands on the host; review what it executes and whether any input reaches it.

MEDIUMdesign/src/cli.ts:34 — Code-execution surface: a node child_process call site. The server can run commands on the host; review what it executes and whether any input reaches it.

MEDIUMdesign/src/cli.ts:185 — Code-execution surface: a exec/spawn call site. The server can run commands on the host; review what it executes and whether any input reaches it.

MEDIUMdesign/src/cli.ts:224 — Code-execution surface: a exec/spawn call site. The server can run commands on the host; review what it executes and whether any input reaches it.

MEDIUMdesign/src/daemon-client.ts:23 — Code-execution surface: a node child_process call site. The server can run commands on the host; review what it executes and whether any input reaches it.

MEDIUMdesign/src/daemon-client.ts:24 — Code-execution surface: a exec/spawn call site. The server can run commands on the host; review what it executes and whether any input reaches it.

+ 902 more in this check

Secrets & credentials93 findings

INFOdesign/prototype.ts:12 — Reads a secret-shaped environment variable. Ordinary for a credentialed server; recorded for completeness.

INFOdesign/test/auth.test.ts:35 — Reads a secret-shaped environment variable. Ordinary for a credentialed server; recorded for completeness.

INFOdesign/test/auth.test.ts:40 — Reads a secret-shaped environment variable. Ordinary for a credentialed server; recorded for completeness.

INFOdesign/test/auth.test.ts:49 — Reads a secret-shaped environment variable. Ordinary for a credentialed server; recorded for completeness.

INFOdesign/test/auth.test.ts:50 — Reads a secret-shaped environment variable. Ordinary for a credentialed server; recorded for completeness.

INFOdesign/test/auth.test.ts:59 — Reads a secret-shaped environment variable. Ordinary for a credentialed server; recorded for completeness.

+ 87 more in this check

Network calls out149 findings

MEDIUMdesign/prototype.ts:47 — Hardcoded external endpoint 'api.openai.com'. STATIC signal only: this flags a declared destination for human or dynamic-egress confirmation; it does NOT assert exfiltration.

MEDIUMdesign/src/auth.ts:130 — Hardcoded external endpoint 'platform.openai.com'. STATIC signal only: this flags a declared destination for human or dynamic-egress confirmation; it does NOT assert exfiltration.

MEDIUMdesign/src/check.ts:25 — Hardcoded external endpoint 'api.openai.com'. STATIC signal only: this flags a declared destination for human or dynamic-egress confirmation; it does NOT assert exfiltration.

MEDIUMdesign/src/check.ts:67 — Hardcoded external endpoint 'platform.openai.com'. STATIC signal only: this flags a declared destination for human or dynamic-egress confirmation; it does NOT assert exfiltration.

MEDIUMdesign/src/cli.ts:88 — Hardcoded external endpoint 'platform.openai.com'. STATIC signal only: this flags a declared destination for human or dynamic-egress confirmation; it does NOT assert exfiltration.

MEDIUMdesign/src/design-to-code.ts:40 — Hardcoded external endpoint 'api.openai.com'. STATIC signal only: this flags a declared destination for human or dynamic-egress confirmation; it does NOT assert exfiltration.

+ 143 more in this check

Prompt-injection passthrough13 findings

INFOdesign/test/serve.test.ts — HEURISTIC: this file both fetches external content and returns content as tool output, with no obvious sanitization. External text returned into tool output can carry instructions an agent obeys (prompt-injection passthrough). Confirm manually; this is a hint, not proof.

INFOdesign/src/check.ts — HEURISTIC: this file both fetches external content and returns content as tool output, with no obvious sanitization. External text returned into tool output can carry instructions an agent obeys (prompt-injection passthrough). Confirm manually; this is a hint, not proof.

INFOdesign/src/design-to-code.ts — HEURISTIC: this file both fetches external content and returns content as tool output, with no obvious sanitization. External text returned into tool output can carry instructions an agent obeys (prompt-injection passthrough). Confirm manually; this is a hint, not proof.

INFOdesign/src/diff.ts — HEURISTIC: this file both fetches external content and returns content as tool output, with no obvious sanitization. External text returned into tool output can carry instructions an agent obeys (prompt-injection passthrough). Confirm manually; this is a hint, not proof.

INFOdesign/src/evolve.ts — HEURISTIC: this file both fetches external content and returns content as tool output, with no obvious sanitization. External text returned into tool output can carry instructions an agent obeys (prompt-injection passthrough). Confirm manually; this is a hint, not proof.

INFOdesign/src/memory.ts — HEURISTIC: this file both fetches external content and returns content as tool output, with no obvious sanitization. External text returned into tool output can carry instructions an agent obeys (prompt-injection passthrough). Confirm manually; this is a hint, not proof.

+ 7 more in this check

Permission scope breadth32 findings

INFOdesign/test/daemon-discovery.test.ts — HEURISTIC: broad capability surface in one file (filesystem, network, subprocess). A scope-breadth hint: the more distinct host capabilities a server touches, the more a buyer is granting. Confirm it matches the stated function.

INFOdesign/test/feedback-roundtrip-daemon.test.ts — HEURISTIC: broad capability surface in one file (filesystem, network, subprocess). A scope-breadth hint: the more distinct host capabilities a server touches, the more a buyer is granting. Confirm it matches the stated function.

INFOdesign/test/serve.test.ts — HEURISTIC: broad capability surface in one file (filesystem, network, subprocess). A scope-breadth hint: the more distinct host capabilities a server touches, the more a buyer is granting. Confirm it matches the stated function.

INFOdesign/src/daemon-client.ts — HEURISTIC: broad capability surface in one file (filesystem, network, subprocess). A scope-breadth hint: the more distinct host capabilities a server touches, the more a buyer is granting. Confirm it matches the stated function.

INFOdesign/src/daemon-state.ts — HEURISTIC: broad capability surface in one file (filesystem, network, subprocess). A scope-breadth hint: the more distinct host capabilities a server touches, the more a buyer is granting. Confirm it matches the stated function.

INFOdesign/src/serve.ts — HEURISTIC: broad capability surface in one file (filesystem, network, subprocess). A scope-breadth hint: the more distinct host capabilities a server touches, the more a buyer is granting. Confirm it matches the stated function.

+ 26 more in this check

How to read this: these are static checks over the source at a point in time. They catch the patterns above, not everything. Absence of a flag is not absence of danger, and a tool that runs cleanly can still behave differently once installed. We do not call any tool simply "safe". Runtime-behavior checks are the next layer we are adding.

04RELATEDWORKS ALONGSIDE THIS

From the same session