[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M72TFB4WD2ESYAHPMAV5E4PS":3,"skill-same:01M72TFB4WD2ESYAHPMAV5E4PS":79},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":18,"linkOnly":19,"markdownContent":20,"markdownBody":21,"frontMatter":22,"refName":23,"commitSha":24,"skillDirectory":25,"forksCount":26,"repositoryHtmlUrl":27,"repositoryHomepageUrl":6,"ownerHtmlUrl":28,"ownerBlogUrl":29,"ownerDescription":6,"updatedAt":30,"install":31,"resources":44},"01M72TFB4WD2ESYAHPMAV5E4PS","ywc-tech-research",null,"(ywc) Compare libraries, implementation approaches and technical trade-offs and sourced best practices before plan\u002Fspec decisions. Triggers: \"기술 조사\", \"라이브러리 비교\", \"research\", \"compare options\", \"best way to implement\", \"技術調査\", \"ライブラリ比較\". For business value or ROI use ywc-product-review. Do not use for spec writing (ywc-spec-writer), validation (ywc-spec-validate), generation (ywc-code-gen), or simple doc lookup without evaluation.","yongwoon\u002Fywc-agent-toolkit","yongwoon","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F9317029?v=4","01M72SZSF6VKANKG5QTMV4V95S","01M72SZSG8HHC3W2CV5P7EJ5AH",8,"MIT","https:\u002F\u002Fgithub.com\u002Fyongwoon\u002Fywc-agent-toolkit\u002Fblob\u002Fmain\u002Fcodex\u002Fskills\u002Fywc-tech-research\u002FSKILL.md",[],[],"2026-11-10 22:39:42",false,"---\nname: ywc-tech-research\ndescription: >-\n  (ywc) Compare libraries, implementation approaches and technical trade-offs and sourced best practices before plan\u002Fspec decisions. Triggers: \"기술 조사\", \"라이브러리 비교\", \"research\", \"compare options\", \"best way to implement\", \"技術調査\", \"ライブラリ比較\". For business value or ROI use ywc-product-review. Do not use for spec writing (ywc-spec-writer), validation (ywc-spec-validate), generation (ywc-code-gen), or simple doc lookup without evaluation.\n---\n\n# ywc-tech-research\n\n**Announce at start:** \"I'm using the ywc-tech-research skill to investigate options and produce a structured comparison.\"\n\nResearch Agent Skill for technical investigation and library comparison.\n\n## Rationalization Defense\n\nWhen tempted to skip a step, check this table first:\n\n| Excuse | Reality |\n|---|---|\n| \"I know the answer from training data, no need to look up docs\" | Training data lags. Use Context7 \u002F WebFetch for up-to-date sources. |\n| \"One option clearly wins, skip the comparison table\" | The user requested research. Always show the trade-off matrix, even if the recommendation is decisive. |\n| \"Existing project stack constraints are obvious\" | Compatibility with the project's existing stack is the hardest dimension to recover later. Always read package.json \u002F pyproject.toml first. |\n| \"Cite from memory, citation links optional\" | Every claim needs a source link or version-specific reference. Unsourced claims are not research. |\n| \"Combine pros\u002Fcons into one paragraph\" | Use a structured trade-off matrix. Mixed prose hides the comparison. |\n| \"Recommend without flagging unknowns\" | If a dimension is unverified (e.g., long-term maintenance, license edge cases), explicitly state it as a gap. |\n\n**Violating the letter of these rules is violating the spirit.** Research without verifiable sources is opinion.\n\n## Arguments\n\nParse `$ARGUMENTS` for:\n\n| Parameter | Format | Example | Description |\n|-----------|--------|---------|-------------|\n| Topic | free text | `\"Hono SSE implementation\"` | Research topic (required) |\n| `--compare` | `--compare \"A,B\"` | `--compare \"Redis,Valkey\"` | Comma-separated list of options to compare in parallel |\n| `--depth` | `--depth 25\\|50\\|75\\|100` | `--depth 50` | Output depth level (default: 50). 25=summary-only; 50=standard; 75=detailed with evidence; 100=exhaustive with full source excerpts and gap analysis |\n| `--output` | `--output \u003Cpath>` | _(optional)_ | Persist the report to a repository-relative Markdown path under `docs\u002Fywc-plans\u002F` unless the project defines another safe root explicitly. |\n| `--overwrite` | flag | off | Request overwrite of an existing persisted report. Never effective by itself. |\n| `--confirm-overwrite` | flag | off | Explicitly confirms overwrite of the same approved output path. Required with `--overwrite`. |\n| `--non-interactive` | flag | off | Allows autonomous overwrite only when `--output`, `--overwrite`, and `--confirm-overwrite` are all present and safe. |\n\n## Safe Persistence Contract\n\nWhen `--output` is supplied, persist only to a repository-relative Markdown path.\nReject absolute paths, `..` traversal, symlink escape, and non-Markdown\nextensions before any write semantics are described. The default safe root is\n`docs\u002Fywc-plans\u002F` unless the project explicitly defines another root.\n\nOverwrite is gated:\n\n- new path: `--output \u003Cpath>` is sufficient\n- existing path: require `--output \u003Cpath> --overwrite --confirm-overwrite`\n- autonomous overwrite: require the same flags plus `--non-interactive`\n\nMissing confirmation, path mismatch, or unsafe destinations return\n`NEEDS_CONTEXT`.\n\n## Execution Steps\n\n1. **Collect Project Context** — Read `CLAUDE.md`, `package.json`, and directory structure to identify the current tech stack and any constraints (runtime, license, bundle size).\n\n2. **Plan Research Strategy** — Determine approach based on topic characteristics:\n   - Library comparison → official docs + package registry stats + GitHub activity\n   - Implementation approach → official docs + code examples + existing project patterns\n   - Architecture decision → technical blogs + comparable case studies + trade-off analysis\n\n3. **Gather Information** — Use sources in this priority order:\n\n   **Step 3a — context7 MCP (required, not optional)**\n\n   Always query context7 before any web search. context7 returns curated, version-pinned documentation that training data cannot guarantee. Skip context7 only if it explicitly returns no results for the topic — in that case, note the fallback in the References section.\n\n   ```\n   mcp__context7__resolve-library-id(libraryName: \"\u003Clibrary>\")\n   mcp__context7__query-docs(context7CompatibleLibraryID: \"\u003Cid>\", topic: \"\u003Ctopic>\")\n   ```\n\n   **Step 3b — Official documentation and package registries**\n\n   Supplement context7 with direct official docs. For any source, check the last-updated date:\n   - Prefer documentation updated within the last 12 months\n   - Flag sources older than 2 years as `[potentially stale]` in the References section\n   - Distinguish official docs from community resources (Stack Overflow, blogs) — present them separately\n\n   **Step 3c — `--compare` parallel mode**\n\n   When `--compare \"A,B[,C...]\"` is specified, spawn one research subagent per option in parallel via Codex subagent tools, each with the same project context and quality criteria. Collect all results before synthesizing — do not synthesize from partial results.\n\n4. **Verify Source Recency** — Before writing the report, confirm:\n   - At least one official source per option was found\n   - Each official source has a visible last-updated date or release version\n   - If only community sources are available, state this explicitly in the report\n\n5. **Analyze and Output Report** — Write a structured report focused on compatibility with the current project stack. If the recommendation still depends on unresolved facts, add them under `### Unknowns Surfaced` and downgrade status accordingly.\n\n## Source Discipline Rules\n\nThese rules apply to every claim in the report, regardless of `--depth` level. They exist to prevent confident-sounding hallucination from polluting technical decisions.\n\n1. **Every factual claim must cite a source** — write `[source: \u003Curl or \"context7\">]` inline after the claim. If no source was found, write `[source: NOT FOUND — inference only]`.\n2. **Distinguish fact from inference** — prefix inferred claims with `[INFERRED]`. Inferences are acceptable, but must be labeled so readers can weigh them appropriately.\n3. **Cross-reference single-source claims** — any claim with only one source and high decision impact (e.g., \"library X has known security issues\") must be confirmed with a second source before being reported as fact. If a second source cannot be found, label it `[SINGLE SOURCE — verify before acting]`.\n4. **Prefer sources ≤12 months old** — sources older than 12 months must be flagged with `[potentially stale: \u003Cdate>]`. Sources older than 2 years should not be cited without an explicit caveat.\n5. **Acknowledge gaps explicitly** — if a required Quality Criteria axis cannot be sourced, write the full `N\u002FA — no data found` entry. Never omit a row silently.\n6. **No gap-filling from training data alone** — Claude's internal knowledge may be used to explain concepts, but version numbers, benchmark figures, and security advisories must come from a verifiable external source fetched in this session.\n\n## Research Quality Criteria\n\nEvery option in a comparative analysis must be assessed on all axes below. If data for an axis is unavailable, write `N\u002FA — no data found` rather than omitting the row.\n\n| Axis | What to Check | Verification Method |\n|------|---------------|---------------------|\n| Functional Fit | Does it meet the stated requirements? | Feature matrix against requirements |\n| Ecosystem Health | Maintenance status, release cadence, open issue count | GitHub Insights, npm\u002FPyPI stats |\n| Learning Curve | Adoption cost relative to the team's current stack | Migration guide existence, API similarity |\n| Performance | Throughput, latency, or bundle size vs. alternatives | Official benchmarks or community benchmark links |\n| License | Commercial use restrictions, copyleft implications | SPDX identifier from package registry |\n| Source Recency | Age of official documentation | Last-updated date or latest release tag |\n\n## Output Format\n\nOutput depth is controlled by `--depth` (default: 50):\n\n| Level | Label | Contents |\n|-------|-------|----------|\n| 25 | Essential | Summary + Recommendation only. No comparison table, no references section. |\n| 50 | Standard | Full report below. References section with top 3 sources per option. |\n| 75 | Detailed | Full report + per-claim `[source: …]` inline citations + `[INFERRED]` labels throughout. |\n| 100 | Exhaustive | Everything in 75 + direct source excerpts (≤3 sentences each) + explicit gap analysis listing every claim that could not be sourced. |\n\n```text\n## Research Result: {topic}\n\n### Summary\n(1-2 sentence key conclusion. Inferences labeled [INFERRED].)\n\n### Comparative Analysis\n| Option | Pros | Cons | Ecosystem Health | License | Source Age | Fit |\n|--------|------|------|-----------------|---------|------------|-----|\n(At --depth 75\u002F100: each cell's key claim includes inline [source: …] citation)\n\n### Recommended Approach\n(Recommendation with rationale, citing specific axes from the table above.\n Any inference prefixed with [INFERRED]. Any single-source claim marked [SINGLE SOURCE — verify before acting].)\n\n### Project-Specific Considerations\n(Compatibility with current stack, migration cost, known conflicts, and whether the result should be preserved in a local discovery artifact)\n\n### Unknowns Surfaced\n(Only unresolved questions or assumptions that materially affect the recommendation. Omit when none.)\n\n### References\n- [Official] {name} — {url} (last updated: {date})\n- [Community] {name} — {url} (flagged [potentially stale: \u003Cdate>] if >12 months old)\n- [context7] used: yes\u002Fno — if no, state the reason\n\n### Persistence Metadata\n- Output Path: {project-relative markdown path}\n- Fetch Date (UTC): {ISO-8601 timestamp}\n- Version \u002F Release Hints: {versions or dates cited}\n- Provenance Notes: {how evidence, [INFERRED], and known gaps were separated}\n\n### Completion Status\n(One of: DONE | DONE_WITH_CONCERNS | BLOCKED | NEEDS_CONTEXT)\n```\n\n**Completion Status rules:**\n\n| Status | When to use |\n|--------|------------|\n| `DONE` | All axes covered with official sources, recommendation is clear |\n| `DONE_WITH_CONCERNS` | Research complete but with caveats — stale sources, missing benchmark data, or a recommendation that still depends on unresolved unknowns |\n| `BLOCKED` | Cannot research the topic — context7 returned nothing and no official docs are publicly accessible |\n| `NEEDS_CONTEXT` | Topic is too vague to produce a useful recommendation without clarification from the user |\n\n## Validation\n\nBefore returning the recommendation, verify that official sources were checked first, source dates are present, stale\u002Fcommunity-only evidence is labeled, unsupported claims are marked `[INFERRED]`, and the final status reflects any missing benchmark or compatibility evidence.\n\n## Integration\n\n- **upstream**: None (standalone execution)\n- **downstream**: Used as reference material when writing specifications or as input to `ywc-task-generator`\n- **multi-session discovery**: When the research leaves multiple unresolved follow-up decisions that must survive across sessions, hand the result to `ywc-wayfinder` as a local map\u002Fticket reference instead of keeping the state only in chat.\n- **persisted handoff**: When a report is saved with `--output`, downstream skills should cite the project-relative artifact path and preserve its overwrite \u002F provenance expectations rather than copying unsourced summaries.\n","# ywc-tech-research\n\n**Announce at start:** \"I'm using the ywc-tech-research skill to investigate options and produce a structured comparison.\"\n\nResearch Agent Skill for technical investigation and library comparison.\n\n## Rationalization Defense\n\nWhen tempted to skip a step, check this table first:\n\n| Excuse | Reality |\n|---|---|\n| \"I know the answer from training data, no need to look up docs\" | Training data lags. Use Context7 \u002F WebFetch for up-to-date sources. |\n| \"One option clearly wins, skip the comparison table\" | The user requested research. Always show the trade-off matrix, even if the recommendation is decisive. |\n| \"Existing project stack constraints are obvious\" | Compatibility with the project's existing stack is the hardest dimension to recover later. Always read package.json \u002F pyproject.toml first. |\n| \"Cite from memory, citation links optional\" | Every claim needs a source link or version-specific reference. Unsourced claims are not research. |\n| \"Combine pros\u002Fcons into one paragraph\" | Use a structured trade-off matrix. Mixed prose hides the comparison. |\n| \"Recommend without flagging unknowns\" | If a dimension is unverified (e.g., long-term maintenance, license edge cases), explicitly state it as a gap. |\n\n**Violating the letter of these rules is violating the spirit.** Research without verifiable sources is opinion.\n\n## Arguments\n\nParse `$ARGUMENTS` for:\n\n| Parameter | Format | Example | Description |\n|-----------|--------|---------|-------------|\n| Topic | free text | `\"Hono SSE implementation\"` | Research topic (required) |\n| `--compare` | `--compare \"A,B\"` | `--compare \"Redis,Valkey\"` | Comma-separated list of options to compare in parallel |\n| `--depth` | `--depth 25\\|50\\|75\\|100` | `--depth 50` | Output depth level (default: 50). 25=summary-only; 50=standard; 75=detailed with evidence; 100=exhaustive with full source excerpts and gap analysis |\n| `--output` | `--output \u003Cpath>` | _(optional)_ | Persist the report to a repository-relative Markdown path under `docs\u002Fywc-plans\u002F` unless the project defines another safe root explicitly. |\n| `--overwrite` | flag | off | Request overwrite of an existing persisted report. Never effective by itself. |\n| `--confirm-overwrite` | flag | off | Explicitly confirms overwrite of the same approved output path. Required with `--overwrite`. |\n| `--non-interactive` | flag | off | Allows autonomous overwrite only when `--output`, `--overwrite`, and `--confirm-overwrite` are all present and safe. |\n\n## Safe Persistence Contract\n\nWhen `--output` is supplied, persist only to a repository-relative Markdown path.\nReject absolute paths, `..` traversal, symlink escape, and non-Markdown\nextensions before any write semantics are described. The default safe root is\n`docs\u002Fywc-plans\u002F` unless the project explicitly defines another root.\n\nOverwrite is gated:\n\n- new path: `--output \u003Cpath>` is sufficient\n- existing path: require `--output \u003Cpath> --overwrite --confirm-overwrite`\n- autonomous overwrite: require the same flags plus `--non-interactive`\n\nMissing confirmation, path mismatch, or unsafe destinations return\n`NEEDS_CONTEXT`.\n\n## Execution Steps\n\n1. **Collect Project Context** — Read `CLAUDE.md`, `package.json`, and directory structure to identify the current tech stack and any constraints (runtime, license, bundle size).\n\n2. **Plan Research Strategy** — Determine approach based on topic characteristics:\n   - Library comparison → official docs + package registry stats + GitHub activity\n   - Implementation approach → official docs + code examples + existing project patterns\n   - Architecture decision → technical blogs + comparable case studies + trade-off analysis\n\n3. **Gather Information** — Use sources in this priority order:\n\n   **Step 3a — context7 MCP (required, not optional)**\n\n   Always query context7 before any web search. context7 returns curated, version-pinned documentation that training data cannot guarantee. Skip context7 only if it explicitly returns no results for the topic — in that case, note the fallback in the References section.\n\n   ```\n   mcp__context7__resolve-library-id(libraryName: \"\u003Clibrary>\")\n   mcp__context7__query-docs(context7CompatibleLibraryID: \"\u003Cid>\", topic: \"\u003Ctopic>\")\n   ```\n\n   **Step 3b — Official documentation and package registries**\n\n   Supplement context7 with direct official docs. For any source, check the last-updated date:\n   - Prefer documentation updated within the last 12 months\n   - Flag sources older than 2 years as `[potentially stale]` in the References section\n   - Distinguish official docs from community resources (Stack Overflow, blogs) — present them separately\n\n   **Step 3c — `--compare` parallel mode**\n\n   When `--compare \"A,B[,C...]\"` is specified, spawn one research subagent per option in parallel via Codex subagent tools, each with the same project context and quality criteria. Collect all results before synthesizing — do not synthesize from partial results.\n\n4. **Verify Source Recency** — Before writing the report, confirm:\n   - At least one official source per option was found\n   - Each official source has a visible last-updated date or release version\n   - If only community sources are available, state this explicitly in the report\n\n5. **Analyze and Output Report** — Write a structured report focused on compatibility with the current project stack. If the recommendation still depends on unresolved facts, add them under `### Unknowns Surfaced` and downgrade status accordingly.\n\n## Source Discipline Rules\n\nThese rules apply to every claim in the report, regardless of `--depth` level. They exist to prevent confident-sounding hallucination from polluting technical decisions.\n\n1. **Every factual claim must cite a source** — write `[source: \u003Curl or \"context7\">]` inline after the claim. If no source was found, write `[source: NOT FOUND — inference only]`.\n2. **Distinguish fact from inference** — prefix inferred claims with `[INFERRED]`. Inferences are acceptable, but must be labeled so readers can weigh them appropriately.\n3. **Cross-reference single-source claims** — any claim with only one source and high decision impact (e.g., \"library X has known security issues\") must be confirmed with a second source before being reported as fact. If a second source cannot be found, label it `[SINGLE SOURCE — verify before acting]`.\n4. **Prefer sources ≤12 months old** — sources older than 12 months must be flagged with `[potentially stale: \u003Cdate>]`. Sources older than 2 years should not be cited without an explicit caveat.\n5. **Acknowledge gaps explicitly** — if a required Quality Criteria axis cannot be sourced, write the full `N\u002FA — no data found` entry. Never omit a row silently.\n6. **No gap-filling from training data alone** — Claude's internal knowledge may be used to explain concepts, but version numbers, benchmark figures, and security advisories must come from a verifiable external source fetched in this session.\n\n## Research Quality Criteria\n\nEvery option in a comparative analysis must be assessed on all axes below. If data for an axis is unavailable, write `N\u002FA — no data found` rather than omitting the row.\n\n| Axis | What to Check | Verification Method |\n|------|---------------|---------------------|\n| Functional Fit | Does it meet the stated requirements? | Feature matrix against requirements |\n| Ecosystem Health | Maintenance status, release cadence, open issue count | GitHub Insights, npm\u002FPyPI stats |\n| Learning Curve | Adoption cost relative to the team's current stack | Migration guide existence, API similarity |\n| Performance | Throughput, latency, or bundle size vs. alternatives | Official benchmarks or community benchmark links |\n| License | Commercial use restrictions, copyleft implications | SPDX identifier from package registry |\n| Source Recency | Age of official documentation | Last-updated date or latest release tag |\n\n## Output Format\n\nOutput depth is controlled by `--depth` (default: 50):\n\n| Level | Label | Contents |\n|-------|-------|----------|\n| 25 | Essential | Summary + Recommendation only. No comparison table, no references section. |\n| 50 | Standard | Full report below. References section with top 3 sources per option. |\n| 75 | Detailed | Full report + per-claim `[source: …]` inline citations + `[INFERRED]` labels throughout. |\n| 100 | Exhaustive | Everything in 75 + direct source excerpts (≤3 sentences each) + explicit gap analysis listing every claim that could not be sourced. |\n\n```text\n## Research Result: {topic}\n\n### Summary\n(1-2 sentence key conclusion. Inferences labeled [INFERRED].)\n\n### Comparative Analysis\n| Option | Pros | Cons | Ecosystem Health | License | Source Age | Fit |\n|--------|------|------|-----------------|---------|------------|-----|\n(At --depth 75\u002F100: each cell's key claim includes inline [source: …] citation)\n\n### Recommended Approach\n(Recommendation with rationale, citing specific axes from the table above.\n Any inference prefixed with [INFERRED]. Any single-source claim marked [SINGLE SOURCE — verify before acting].)\n\n### Project-Specific Considerations\n(Compatibility with current stack, migration cost, known conflicts, and whether the result should be preserved in a local discovery artifact)\n\n### Unknowns Surfaced\n(Only unresolved questions or assumptions that materially affect the recommendation. Omit when none.)\n\n### References\n- [Official] {name} — {url} (last updated: {date})\n- [Community] {name} — {url} (flagged [potentially stale: \u003Cdate>] if >12 months old)\n- [context7] used: yes\u002Fno — if no, state the reason\n\n### Persistence Metadata\n- Output Path: {project-relative markdown path}\n- Fetch Date (UTC): {ISO-8601 timestamp}\n- Version \u002F Release Hints: {versions or dates cited}\n- Provenance Notes: {how evidence, [INFERRED], and known gaps were separated}\n\n### Completion Status\n(One of: DONE | DONE_WITH_CONCERNS | BLOCKED | NEEDS_CONTEXT)\n```\n\n**Completion Status rules:**\n\n| Status | When to use |\n|--------|------------|\n| `DONE` | All axes covered with official sources, recommendation is clear |\n| `DONE_WITH_CONCERNS` | Research complete but with caveats — stale sources, missing benchmark data, or a recommendation that still depends on unresolved unknowns |\n| `BLOCKED` | Cannot research the topic — context7 returned nothing and no official docs are publicly accessible |\n| `NEEDS_CONTEXT` | Topic is too vague to produce a useful recommendation without clarification from the user |\n\n## Validation\n\nBefore returning the recommendation, verify that official sources were checked first, source dates are present, stale\u002Fcommunity-only evidence is labeled, unsupported claims are marked `[INFERRED]`, and the final status reflects any missing benchmark or compatibility evidence.\n\n## Integration\n\n- **upstream**: None (standalone execution)\n- **downstream**: Used as reference material when writing specifications or as input to `ywc-task-generator`\n- **multi-session discovery**: When the research leaves multiple unresolved follow-up decisions that must survive across sessions, hand the result to `ywc-wayfinder` as a local map\u002Fticket reference instead of keeping the state only in chat.\n- **persisted handoff**: When a report is saved with `--output`, downstream skills should cite the project-relative artifact path and preserve its overwrite \u002F provenance expectations rather than copying unsourced summaries.\n",{"name":5,"description":7},"main","7e8ac872e53df8c89d75663e031626281a2a2f5c","codex\u002Fskills\u002Fywc-tech-research",1,"https:\u002F\u002Fgithub.com\u002Fyongwoon\u002Fywc-agent-toolkit","https:\u002F\u002Fgithub.com\u002Fyongwoon","https:\u002F\u002Fwww.webhani.com","2026-11-08 17:14:59",{"source":8,"skillName":5,"commands":32,"prompt":42,"zipAvailable":43},[33,36,39],{"tool":34,"command":35},"npx","npx skills add yongwoon\u002Fywc-agent-toolkit --skill ywc-tech-research",{"tool":37,"command":38},"bunx","bunx skills add yongwoon\u002Fywc-agent-toolkit --skill ywc-tech-research",{"tool":40,"command":41},"pnpm","pnpm dlx skills add yongwoon\u002Fywc-agent-toolkit --skill ywc-tech-research","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fyongwoon\u002Fywc-agent-toolkit\u002Fblob\u002Fmain\u002Fcodex\u002Fskills\u002Fywc-tech-research\u002FSKILL.md",true,[45,51,55,59,63,67,71,75],{"path":46,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":49,"sourceHtmlUrl":50},"agents\u002Fopenai.yaml","REFERENCE","参考文档",543,"https:\u002F\u002Fgithub.com\u002Fyongwoon\u002Fywc-agent-toolkit\u002Fblob\u002Fmain\u002Fcodex\u002Fskills\u002Fywc-tech-research\u002Fagents\u002Fopenai.yaml",{"path":52,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":53,"sourceHtmlUrl":54},"evals\u002Fevals.json",2278,"https:\u002F\u002Fgithub.com\u002Fyongwoon\u002Fywc-agent-toolkit\u002Fblob\u002Fmain\u002Fcodex\u002Fskills\u002Fywc-tech-research\u002Fevals\u002Fevals.json",{"path":56,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":57,"sourceHtmlUrl":58},"README.en.md",1390,"https:\u002F\u002Fgithub.com\u002Fyongwoon\u002Fywc-agent-toolkit\u002Fblob\u002Fmain\u002Fcodex\u002Fskills\u002Fywc-tech-research\u002FREADME.en.md",{"path":60,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":61,"sourceHtmlUrl":62},"README.es.md",1514,"https:\u002F\u002Fgithub.com\u002Fyongwoon\u002Fywc-agent-toolkit\u002Fblob\u002Fmain\u002Fcodex\u002Fskills\u002Fywc-tech-research\u002FREADME.es.md",{"path":64,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":65,"sourceHtmlUrl":66},"README.ja.md",1144,"https:\u002F\u002Fgithub.com\u002Fyongwoon\u002Fywc-agent-toolkit\u002Fblob\u002Fmain\u002Fcodex\u002Fskills\u002Fywc-tech-research\u002FREADME.ja.md",{"path":68,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":69,"sourceHtmlUrl":70},"README.ko.md",1081,"https:\u002F\u002Fgithub.com\u002Fyongwoon\u002Fywc-agent-toolkit\u002Fblob\u002Fmain\u002Fcodex\u002Fskills\u002Fywc-tech-research\u002FREADME.ko.md",{"path":72,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":73,"sourceHtmlUrl":74},"README.md",1126,"https:\u002F\u002Fgithub.com\u002Fyongwoon\u002Fywc-agent-toolkit\u002Fblob\u002Fmain\u002Fcodex\u002Fskills\u002Fywc-tech-research\u002FREADME.md",{"path":76,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":77,"sourceHtmlUrl":78},"README.zh.md",1289,"https:\u002F\u002Fgithub.com\u002Fyongwoon\u002Fywc-agent-toolkit\u002Fblob\u002Fmain\u002Fcodex\u002Fskills\u002Fywc-tech-research\u002FREADME.zh.md",[80,86,92,99,106,113],{"skillNo":81,"name":82,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":83,"categories":84,"keywords":85,"publishedAt":18},"01M72TQ84GTCPCV2VJ778V17TW","missing-frontmatter","https:\u002F\u002Fgithub.com\u002Fyongwoon\u002Fywc-agent-toolkit\u002Fblob\u002Fmain\u002Fscripts\u002Ffixtures\u002Fcodex-description-limit\u002Fmissing-frontmatter\u002FSKILL.md",[],[],{"skillNo":87,"name":88,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":89,"categories":90,"keywords":91,"publishedAt":18},"01M72TQ6M4BFWE7P9EPW0GZDCM","malformed-frontmatter","https:\u002F\u002Fgithub.com\u002Fyongwoon\u002Fywc-agent-toolkit\u002Fblob\u002Fmain\u002Fscripts\u002Ffixtures\u002Fcodex-description-limit\u002Fmalformed-frontmatter\u002FSKILL.md",[],[],{"skillNo":93,"name":94,"nameZh":6,"description":95,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":96,"categories":97,"keywords":98,"publishedAt":18},"01M72TQ3ESGKXZ4QRWGTX92BKJ","fixture-501-fail","(ywc) Use when aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa Do not use for x.","https:\u002F\u002Fgithub.com\u002Fyongwoon\u002Fywc-agent-toolkit\u002Fblob\u002Fmain\u002Fscripts\u002Ffixtures\u002Fcodex-description-limit\u002F501-fail\u002FSKILL.md",[],[],{"skillNo":100,"name":101,"nameZh":6,"description":102,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":103,"categories":104,"keywords":105,"publishedAt":18},"01M72TQ0HJ8H3R37J9WPWGKJZR","fixture-500-pass","(ywc) Use when aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa Do not use for x.","https:\u002F\u002Fgithub.com\u002Fyongwoon\u002Fywc-agent-toolkit\u002Fblob\u002Fmain\u002Fscripts\u002Ffixtures\u002Fcodex-description-limit\u002F500-pass\u002FSKILL.md",[],[],{"skillNo":107,"name":108,"nameZh":6,"description":109,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":110,"categories":111,"keywords":112,"publishedAt":18},"01M72TPWNCC7NEYD0D9N12KSQS","ywc-worktrees","(ywc) Use when creating, auditing, pruning, or resolving git worktrees for isolated task workspaces, whether delegated from an executor or used standalone. Triggers: \"worktree 생성\", \"worktree audit\", \"worktree 정리\", \"worktree prune\", \"ywc-worktrees\", \"create worktree\". Do not use for branch creation\u002Fdeletion alone, PR creation (use ywc-create-pr), CI or merge delivery (use ywc-finish-branch), task orchestration (use ywc-sequential-executor or ywc-parallel-executor), or non-git isolation.","https:\u002F\u002Fgithub.com\u002Fyongwoon\u002Fywc-agent-toolkit\u002Fblob\u002Fmain\u002Fplugins\u002Fywc-agent-toolkit\u002Fskills\u002Fywc-worktrees\u002FSKILL.md",[],[],{"skillNo":114,"name":115,"nameZh":6,"description":116,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":117,"categories":118,"keywords":119,"publishedAt":18},"01M72TPRNR3E6QWB91GM98ABBV","ywc-wayfinder","(ywc) Use when a large or uncertain change needs multi-session discovery with a deterministic local map and exactly one active ticket before planning or implementation. Triggers: \"wayfinder\", \"길찾기\", \"discovery map\", \"multi-session discovery\", \"탐색 ticket\", \"経路整理\", \"探索マップ\", \"ywc-wayfinder\". Do not use for ordinary planning (use ywc-plan), idea clarification (use ywc-brainstorm), direct implementation, or tracker\u002Fdatabase writes outside the local Markdown map.","https:\u002F\u002Fgithub.com\u002Fyongwoon\u002Fywc-agent-toolkit\u002Fblob\u002Fmain\u002Fplugins\u002Fywc-agent-toolkit\u002Fskills\u002Fywc-wayfinder\u002FSKILL.md",[],[]]