Windy3f3f3f3f/claude-code-from-scratch
Build your own Claude Code from scratch. 🔍 Claude Code 开源了 50 万行代码，读不动？用 ~5000 行 TypeScript / Python 从零复现核心架构，11 章分步教程带你理解 coding agent 精髓
- Repository: https://github.com/Windy3f3f3f3f/claude-code-from-scratch
- Commit:
acc4c34f062b9fbc289ef9cc2be339ba4e970347 - Generated: 2026-11-14T00:15:30+00:00
- Scorer 0.1.0, scoring prov-4+qual-2+sec-2, report schema v2
Scores
| Pillar | Score | Signal | Confidence | Coverage |
|---|---|---|---|---|
| AI provenance (estimated AI-generated share) | 0 | possible | 64% | 100% |
| Code quality | 52 | weak | 90% | 100% |
| Security | 62 | weak | 85% | 100% |
AI provenance
Weak signals consistent with AI-assisted development.
-
Explicitly attributed share (lower bound): 0.0%
-
Unattributed share in AI-consistent velocity bursts: 46.5%
-
Estimated AI-generated share: 0.0%
-
AI tooling configured in repo: claude-code
-
Stylometry index: 0.88 per 1,000 lines (narrated steps 4, placeholders 0, assistant phrasing 0, emoji 16). Reported as a signal, not scored.
| Month | Lines added | Attributed to AI |
|---|---|---|
| 2026-03 | 1,386 | 0% |
| 2026-04 | 13,621 | 0% |
| 2026-05 | 32 | 0% |
| 2026-07 | 7,916 | 0% |
| 2026-10 | 145 | 0% |
Evidence
- .claude/agents/reviewer.md (claude-code configuration) (.claude/agen)
- .claude/rules/chinese-greeting.md (claude-code configuration) (.claude/rule)
- .claude/skills/commit/SKILL.md (claude-code configuration) (.claude/skil)
- .claude/skills/greet/SKILL.md (claude-code configuration) (.claude/skil)
- CLAUDE.md (claude-code configuration) (CLAUDE.md)
- 5,220 lines of first-party code in ~4.2h (1,244 lines/hour, 2 commit session) (a485299eb5c8)
- 3,363 lines of first-party code in ~4.0h (841 lines/hour, 1 commit session) (a51d90943014)
- 2,167 lines of first-party code in ~4.0h (542 lines/hour, 1 commit session) (3abb94529476)
- step-by-step narration comment: // Step 1: Deny rules always win — even bypassPermissions (--yolo) is (src/tools.ts)
- step-by-step narration comment: // Step 2: Plan mode's read-only contract beats allow rules and bypass: (src/tools.ts)
- step-by-step narration comment: // Step 3: Mode-specific logic (src/tools.ts)
- step-by-step narration comment: // Step 3: Built-in dangerous pattern checks (src/tools.ts)
AI provenance describes how code was produced; it is not a quality judgment. Unless attribution is explicit, these are signals consistent with AI generation, not proof. Squash/rebase workflows, imported code, and rewritten history reduce accuracy.
Code quality
Code quality is weak: test-to-source ratio 0.14, no CI test run detected; 25% of function code in complex or long functions; 5% duplicated lines. Weakest area: type safety.
| Area | Score |
|---|---|
| Complexity (lizard) | 47 |
| Duplication | 76 |
| Code smells (Opengrep) | 67 |
| Tests and CI | 31 |
| Documentation | 65 |
| Type safety | 30 |
| Process (Scorecard review/maintenance) | n/a |
- Functions: 749; mean complexity 4.2, 90th percentile 10; 30 complex and 6 long functions
- Duplicated lines: 4.7% of 10,522 compared
- Tests: 1,970 test lines for 13,707 source lines (ratio 0.14); CI test commands: none detected; 9% of recent source commits also change tests
- Docs missing or thin: contributing, changelog
Most complex functions:
- runScenario in steps/test.mjs: complexity 44, 55 lines
- (anonymous) in src/cli.ts: complexity 41, 126 lines
- _chat_openai in python/mini_claude/agent.py: complexity 40, 104 lines
- chatOpenAI in src/agent.ts: complexity 40, 117 lines
- run_repl in python/mini_claude/main.py: complexity 38, 102 lines
Security
Security posture is weak: 9 known vulnerabilities in 3 dependencies; 11 high-severity SAST findings. Weakest area: posture.
| Area | Score |
|---|---|
| Secrets (gitleaks, tree + history) | 97 |
| Static analysis (Opengrep) | 32 |
| Vulnerable dependencies (OSV) | 79 |
| CI workflows (zizmor) | n/a |
| Posture (policy, updates, signing, lockfiles) | 0 |
| OpenSSF Scorecard | n/a |
-
Known vulnerabilities: 9 in 3 packages (critical 0, high 5, medium 4, low 0)
-
Posture checks passed: lockfiles; missing: security policy, dependency updates, signed commits
Top findings
| Severity | Finding | Location |
|---|---|---|
| high | opengrep/python.lang.security.audit.dangerous-asyncio-create-exec-audit: Detected 'create_subprocess_exec' function without a static string. If this data can be controlled by a malicious actor, it may be an instance of command injection. Audit the use of this call to ensure it is not controllable by an external resource. You may consider using 'shlex.escape()'. [confidence low] |
python/mini_claude/mcp_client.py:41 |
| high | opengrep/python.lang.security.audit.dangerous-asyncio-create-exec-tainted-env-args: Detected 'create_subprocess_exec' function with user controlled data. You may consider using 'shlex.escape()'. [confidence low] |
python/mini_claude/mcp_client.py:41 |
| high | opengrep/python.lang.security.audit.dangerous-subprocess-use-audit: Detected subprocess function 'run' without a static string. If this data can be controlled by a malicious actor, it may be an instance of command injection. Audit the use of this call to ensure it is not controllable by an external resource. You may consider using 'shlex.escape()'. [confidence low] |
python/mini_claude/tools.py:356 |
| high | opengrep/python.lang.security.audit.dangerous-subprocess-use-audit: Detected subprocess function 'run' without a static string. If this data can be controlled by a malicious actor, it may be an instance of command injection. Audit the use of this call to ensure it is not controllable by an external resource. You may consider using 'shlex.escape()'. [confidence low] |
python/mini_claude/tools.py:428 |
| high | opengrep/python.lang.security.audit.subprocess-shell-true: Found 'subprocess' function 'run' with 'shell=True'. This is dangerous because this call will spawn the command using a shell process. Doing so propagates current shell settings and variables, which makes it much easier for a malicious actor to execute commands. Use 'shell=False' instead. [confidence medium] |
python/mini_claude/tools.py:430 |
| high | opengrep/javascript.lang.security.detect-child-process: Detected calls to child_process from a function argument input. This could lead to a command injection if the input is user controllable. Try to avoid calls to child_process, and if it is needed ensure user input is correctly sanitized or sandboxed. [confidence low] |
src/tools.ts:500 |
| high | opengrep/python.lang.security.audit.dangerous-subprocess-use-audit: Detected subprocess function 'Popen' without a static string. If this data can be controlled by a malicious actor, it may be an instance of command injection. Audit the use of this call to ensure it is not controllable by an external resource. You may consider using 'shlex.escape()'. [confidence low] |
steps/canonical/py/mcp_client.py:15 |
| high | opengrep/python.lang.security.audit.dangerous-subprocess-use-audit: Detected subprocess function 'run' without a static string. If this data can be controlled by a malicious actor, it may be an instance of command injection. Audit the use of this call to ensure it is not controllable by an external resource. You may consider using 'shlex.escape()'. [confidence low] |
steps/canonical/py/tools.py:209 |
| high | opengrep/python.lang.security.audit.subprocess-shell-true: Found 'subprocess' function 'run' with 'shell=True'. This is dangerous because this call will spawn the command using a shell process. Doing so propagates current shell settings and variables, which makes it much easier for a malicious actor to execute commands. Use 'shell=False' instead. [confidence medium] |
steps/canonical/py/tools.py:209 |
| high | opengrep/javascript.lang.security.detect-child-process: Detected calls to child_process from a function argument command. This could lead to a command injection if the input is user controllable. Try to avoid calls to child_process, and if it is needed ensure user input is correctly sanitized or sandboxed. [confidence low] |
steps/canonical/ts/mcp.ts:14 |
| high | opengrep/javascript.lang.security.detect-child-process: Detected calls to child_process from a function argument input. This could lead to a command injection if the input is user controllable. Try to avoid calls to child_process, and if it is needed ensure user input is correctly sanitized or sandboxed. [confidence low] |
steps/canonical/ts/tools.ts:209 |
| high | osv/GHSA-3jxr-9vmj-r5cp: [email protected] (npm) is affected by GHSA-3jxr-9vmj-r5cp |
package-lock.json |
| high | osv/GHSA-6j4f-fj2g-mc7p: [email protected] (npm) is affected by GHSA-6j4f-fj2g-mc7p |
package-lock.json |
| high | osv/GHSA-mh99-v99m-4gvg: [email protected] (npm) is affected by GHSA-mh99-v99m-4gvg |
package-lock.json |
| high | osv/GHSA-qhr7-859c-m2p7: [email protected] (npm) is affected by GHSA-qhr7-859c-m2p7 |
package-lock.json |
| high | osv/GHSA-rgw5-rvv9-x895: [email protected] (npm) is affected by GHSA-rgw5-rvv9-x895 |
package-lock.json |
| medium | opengrep/python.lang.security.audit.dynamic-urllib-use-detected: Detected a dynamic value being used with urllib. urllib supports 'file://' schemes, so a dynamic value controlled by a malicious actor may allow them to read arbitrary files. Audit uses of urllib calls to ensure user data cannot control the URLs, or consider using the 'requests' library instead. [confidence low] |
python/mini_claude/tools.py:460 |
| medium | opengrep/javascript.lang.security.audit.path-traversal.path-join-resolve-traversal: Detected possible user input going into a path.join or path.resolve function. This could possibly lead to a path traversal vulnerability, where the attacker can access arbitrary files stored in the file system. Instead, be sure to sanitize or validate user input first. [confidence low] |
src/agent.ts:1324 |
| medium | opengrep/javascript.lang.security.audit.detect-non-literal-fs-filename: Detected that function argument toolName has entered the fs module. An attacker could potentially control the location of this file, to include going backwards in the directory with '../'. To address this, ensure that user-controlled variables in file paths are validated. [confidence low] |
src/agent.ts:1325 |
| medium | opengrep/javascript.lang.security.html-in-template-string: This template literal looks like HTML and has interpolated variables. These variables are not HTML-encoded by default. If the variables contain HTML tags, these may be interpreted by the browser, resulting in cross-site scripting (XSS). [confidence low] |
src/autonomy.ts:461 |
| medium | opengrep/javascript.lang.security.audit.detect-non-literal-fs-filename: Detected that function argument filePath has entered the fs module. An attacker could potentially control the location of this file, to include going backwards in the directory with '../'. To address this, ensure that user-controlled variables in file paths are validated. [confidence low] |
src/mcp.ts:260 |
| medium | opengrep/javascript.lang.security.audit.detect-non-literal-fs-filename: Detected that function argument filePath has entered the fs module. An attacker could potentially control the location of this file, to include going backwards in the directory with '../'. To address this, ensure that user-controlled variables in file paths are validated. [confidence low] |
src/mcp.ts:262 |
| medium | opengrep/javascript.lang.security.audit.detect-non-literal-fs-filename: Detected that function argument a has entered the fs module. An attacker could potentially control the location of this file, to include going backwards in the directory with '../'. To address this, ensure that user-controlled variables in file paths are validated. [confidence low] |
src/memory.ts:83 |
| medium | opengrep/javascript.lang.security.audit.path-traversal.path-join-resolve-traversal: Detected possible user input going into a path.join or path.resolve function. This could possibly lead to a path traversal vulnerability, where the attacker can access arbitrary files stored in the file system. Instead, be sure to sanitize or validate user input first. [confidence low] |
src/memory.ts:83 |
| medium | opengrep/javascript.lang.security.audit.detect-non-literal-fs-filename: Detected that function argument b has entered the fs module. An attacker could potentially control the location of this file, to include going backwards in the directory with '../'. To address this, ensure that user-controlled variables in file paths are validated. [confidence low] |
src/memory.ts:84 |
Static signals only: the scorer never runs the project's code. Secret findings are pattern matches that were not verified, and secret values are never stored.
Repository profile
- Source lines: 13,707; test lines: 1,970
| Language | Lines |
|---|---|
| Python | 6,649 |
| TypeScript | 6,498 |
| JavaScript | 2,421 |
| Shell | 109 |