[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7A028BV2N6CEHKT13B7AEBS":3,"skill-same:01M7A028BV2N6CEHKT13B7AEBS":48},{"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":24,"commitSha":25,"skillDirectory":26,"forksCount":27,"repositoryHtmlUrl":28,"repositoryHomepageUrl":29,"ownerHtmlUrl":30,"ownerBlogUrl":31,"ownerDescription":32,"updatedAt":33,"install":34,"resources":47},"01M7A028BV2N6CEHKT13B7AEBS","Ticket Specs",null,"Create, regenerate, update, and review Xyne Spaces ticket Specifications by interviewing for requirement intent before drafting or writing the Specification.","juspay\u002Fxyne-spaces","juspay","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F11497632?v=4","01M79ZZZBS47MDBJST0J4R2NA0","Juspay Technologies","01M79ZZZCGHA6P27FZG2PJ2P49",827,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Fjuspay\u002Fxyne-spaces\u002Fblob\u002Fmain\u002Fapps\u002Fxyne-claw\u002Fspec-skills\u002Fspec-task\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: Ticket Specs\ndescription: Create, regenerate, update, and review Xyne Spaces ticket Specifications by interviewing for requirement intent before drafting or writing the Specification.\n---\n\n# Xyne Spaces Ticket Specification\n\nUse this skill when creating, generating, regenerating, or updating a Xyne Spaces ticket specification.\n\nThe purpose of the Specification is to capture the requirement intent so that:\n\n- a developer can understand what is expected,\n- a PR reviewer can understand what was requested,\n- and the implementation can later be reviewed against that requirement.\n\nThe Specification must come from requirement context and user clarification.\n\nDo not generate requirement intent merely by describing the implementation.\n\n---\n\n## Specification sections\n\nUse these headings exactly.\n\n### Required\n\n#### Problem statement\n\nDescribe what is wrong, missing, or needed and who or what is affected.\n\nThe Problem statement should explain why this work exists.\n\n#### Solutioning\n\nDescribe the expected functional behaviour after the change is shipped.\n\nFocus on what must be true functionally, not how the developer should implement it.\n\n#### Test cases\n\nDescribe the observable scenarios that confirm the requirement works.\n\nTest cases should make it possible for a developer or reviewer to understand what behaviours need to succeed.\n\n### Optional\n\n#### Implementation details\n\nTechnical details, constraints, or implementation context that are useful to the developer or reviewer.\n\nDo not require this section when there is nothing meaningful to record.\n\n#### Out of scope\n\nAnything explicitly excluded from this ticket.\n\nDo not invent non-goals simply to populate this section.\n\n---\n\n# Before the interview\n\nBefore creating or regenerating a Specification, first understand the requirement context.\n\nIf this is an existing ticket:\n\n1. Fetch the ticket using `spaces-tickets`.\n2. Read:\n   - ticket title,\n   - ticket type, when available,\n   - existing description,\n   - existing Specification, when present,\n   - other relevant requirement context available with the ticket.\n3. Use this information to understand what the ticket is about.\n4. Use it to formulate better and more relevant clarification questions.\n5. Do NOT create or update the Specification yet.\n\nThe existing ticket description is context.\n\nIt does NOT automatically replace the requirement interview.\n\nDo not silently convert an existing one-line description into a complete Specification.\n\nDo not treat the implementation, PR description, commit messages, or code diff as the source of truth for requirement intent.\n\n---\n\n# Mandatory contextual interview\n\nBefore creating or regenerating a Specification, interview the user.\n\nThe Specification sections are fixed:\n\nREQUIRED\n- Problem statement\n- Solutioning\n- Test cases\n\nOPTIONAL\n- Implementation details\n- Out of scope\n\nHowever, the interview questions are NOT fixed.\n\nThe questions MUST be adapted to the context of the specific ticket.\n\nDo not mechanically ask the same questions for every ticket.\n\n---\n\n## Generate questions from the ticket context\n\nUse the ticket title, ticket type, description, current conversation, and other available requirement context to determine what needs clarification.\n\nGenerate questions that help you confidently populate the Specification.\n\nPrioritise questions that clarify:\n\n- what was originally requested,\n- what problem needs to be solved,\n- who or what is affected,\n- expected user-visible behaviour,\n- important product decisions,\n- edge cases,\n- failure behaviour,\n- success criteria,\n- expected test scenarios,\n- technical constraints when relevant,\n- explicit non-goals or boundaries.\n\nDo NOT simply turn the section headings into generic questions.\n\nBad:\n\n- What is the problem statement?\n- What is the solutioning?\n- What are the test cases?\n- What are the implementation details?\n- What is out of scope?\n\nInstead, understand the requirement first and ask concrete questions about the specific feature, bug, or task.\n\n---\n\n## Example — bug\n\nTicket:\n\n> Mention picker shows no users in sandbox.\n\nUseful contextual questions might include:\n\n- When the normal user lookup fails or returns no users, what should the picker show instead?\n- What should the user see if the fallback also fails?\n- Under which environments or conditions is this expected to work?\n- Which scenarios should we verify before considering the bug fixed?\n- Is anything related to mention behaviour explicitly outside this fix?\n\nThese are examples only.\n\nDo not reuse these questions when they are irrelevant to the actual ticket.\n\n---\n\n## Example — feature\n\nTicket:\n\n> Add people to an existing conversation.\n\nUseful contextual questions might include:\n\n- Who should be allowed to add another person to a conversation?\n- What conversation history should a newly added person be able to access?\n- How should files and attachments behave for the newly added person?\n- What should happen if adding the participant fails?\n- What scenarios must work for this feature to be considered complete?\n- Is anything explicitly outside the scope of this feature?\n\nThese are examples only.\n\nGenerate questions appropriate to the actual feature.\n\n---\n\n## Example — technical task\n\nTicket:\n\n> Move ticket specification merging to the Spaces backend.\n\nUseful contextual questions might include:\n\n- What problem with the current specification update flow are we trying to eliminate?\n- What should happen when the ticket already contains a Specification?\n- What existing ticket-description content must be preserved?\n- What should happen if the Specification is malformed or duplicated?\n- What cases should be tested to prove that existing rich-text content is preserved?\n- Are any creation or update paths explicitly outside this change?\n\nThese are examples only.\n\n---\n\n# Question selection\n\nAsk only questions that materially improve the Specification.\n\nDo not ask the user to repeat information they have already explicitly provided.\n\nFor example, if the user has already said:\n\n> Newly added users should see previous messages but should not get access to old attachments.\n\ndo not ask:\n\n> Should newly added users see previous messages?\n\nThat decision is already established.\n\nInstead, identify what remains unclear.\n\nFor example:\n\n- Who is allowed to add people?\n- What happens if adding someone fails?\n- Does the same behaviour apply to private conversations?\n- What scenarios must be verified?\n\n---\n\n## Number of questions\n\nThe number of questions should vary based on the complexity and ambiguity of the ticket.\n\nA simple ticket may require only 2–3 clarification questions.\n\nA complex feature may require 6–10 questions.\n\nDo not ask additional questions merely to reach a fixed number.\n\nAsk the minimum useful set of questions needed to establish the required Specification.\n\n---\n\n## Ask questions in one batch\n\nAsk the initial clarification questions together.\n\nDo not unnecessarily ask one question, wait for an answer, then ask the next question.\n\nThe user should be able to answer the main interview in one response.\n\nAfter asking the questions:\n\nSTOP.\n\nWait for the user's response.\n\nDo NOT create or update the Specification in the same turn in which you ask the interview questions.\n\nDo NOT call `spaces-update-ticket` yet.\n\n---\n\n# Requirement source\n\nProblem statement, Solutioning, and Test cases must represent what was requested or explicitly clarified by the user.\n\nDo not construct the requirement by simply summarising:\n\n- the implementation,\n- PR diff,\n- changed files,\n- commit messages,\n- PR description,\n- or completed code.\n\nGood:\n\n> When the normal user lookup fails, what should the picker show instead?\n\nBad:\n\n> I see that you implemented a fallback to channel participants. Should the requirement say that?\n\nThe implementation may be used as additional context for asking sharper questions.\n\nIt must not become the source of truth for requirement intent.\n\nA Specification derived only from the implementation will naturally resemble the implementation and therefore cannot reliably reveal something that was requested but never implemented.\n\n---\n\n# After the user answers\n\nAfter receiving the user's answers:\n\n1. Map the answers into:\n   - Problem statement,\n   - Solutioning,\n   - Test cases,\n   - Implementation details, when applicable,\n   - Out of scope, when applicable.\n\n2. Use the user's answers and explicitly provided requirement context as the source of truth.\n\n3. Check whether all required sections can now be populated meaningfully.\n\nRequired:\n- Problem statement\n- Solutioning\n- Test cases\n\n4. Do not invent missing information.\n\n5. If required information remains missing or ambiguous, ask only the necessary follow-up questions.\n\n6. Do not repeat the entire interview.\n\n7. Optional sections may be omitted when they are not applicable or the user has provided no meaningful information for them.\n\n---\n\n# Missing information\n\nNever invent a plausible answer.\n\nFor required sections:\n\n- If the user's answer is unclear, ask for clarification.\n- If an important requirement decision remains unresolved, ask about it.\n- Do not consider the Specification complete until the required information is sufficiently clear.\n\nFor optional sections:\n\n- Omit the section if there is no meaningful information.\n- Do not generate filler merely so every heading appears.\n\nDo not use \"Not specified\" to bypass missing required information when the user is available to clarify it.\n\n---\n\n# Draft before updating\n\nOnce sufficient information has been collected, draft the complete Specification.\n\nShow the draft to the user before modifying the ticket.\n\nUse:\n\n## Specification\n\n### Problem statement\n\u003Cdraft>\n\n### Solutioning\n\u003Cdraft>\n\n### Test cases\n\u003Cdraft>\n\n### Implementation details\n\u003Cdraft, only when applicable>\n\n### Out of scope\n\u003Cdraft, only when applicable>\n\nThen ask the user to confirm or correct the Specification.\n\nDo NOT update the ticket until the user confirms the draft.\n\n---\n\n# Confirmation\n\nThe user must have an opportunity to review the generated Specification before it is written to the ticket.\n\nIf the user requests changes:\n\n1. Update the draft.\n2. Show the changed Specification when necessary.\n3. Obtain confirmation.\n\nAfter the user confirms, create or update the ticket Specification.\n\n---\n\n# Existing Specification edits\n\nA full interview is required when:\n\n- creating the first Specification,\n- regenerating the Specification,\n- substantially rewriting the Specification,\n- or the existing Specification does not contain enough requirement information.\n\nA full interview is NOT required for an explicit targeted edit.\n\nExample:\n\n> Add \"notifications are not part of this work\" to Out of scope.\n\nIn that case:\n\n1. Fetch the existing ticket.\n2. Preserve the existing Specification.\n3. Apply only the requested change.\n4. Show the resulting change\u002Fspecification to the user when appropriate.\n5. Update after confirmation.\n\nDo not ask the user to answer Problem statement, Solutioning, and Test cases again for a simple targeted edit.\n\nIf the user says:\n\n- redo the spec,\n- regenerate the spec,\n- rewrite the spec,\n- create a new spec,\n\nrun the contextual interview again.\n\n---\n\n# Pasted requirement or external-agent content\n\nIf the user pastes:\n\n- a requirement,\n- PRD,\n- product discussion,\n- Slack\u002Fthread conversation,\n- notes,\n- output from another coding agent,\n- or other requirement material,\n\nuse it as context.\n\n1. Extract information that is explicitly established.\n2. Map what is already known to the Specification sections.\n3. Identify important ambiguities, missing decisions, edge cases, and missing required information.\n4. Generate contextual clarification questions for those gaps.\n5. Do not ask the user to repeat information already explicitly provided.\n6. Do not silently convert implementation-derived observations into requirement intent.\n7. Wait for the user's answers.\n8. Map the confirmed information into the Specification.\n9. Follow the normal draft → confirmation → update flow.\n\n---\n\n# Writing style\n\nWrite Specifications for developers and PR reviewers.\n\nBe concise.\n\nBe concrete.\n\nPrefer observable behaviour.\n\nAvoid vague requirements.\n\nBad:\n\n> Improve the mention picker.\n\nGood:\n\n> When the primary user lookup returns no users, the mention picker should fall back to eligible channel participants.\n\nAvoid unnecessary implementation prescriptions in Problem statement and Solutioning.\n\nProblem statement should explain the problem.\n\nSolutioning should explain the expected functional outcome.\n\nImplementation details should contain technical implementation constraints when those details actually matter.\n\n---\n\n# Test cases\n\nTest cases should describe meaningful scenarios.\n\nBad:\n\n> Test the feature.\n\nBetter:\n\n> When the primary user lookup returns users, the picker displays those users normally.\n\n> When the primary lookup returns no users, the picker displays eligible channel participants.\n\n> When both the primary lookup and fallback fail, the user sees the specified error state.\n\nInclude positive, failure, or edge scenarios when they were established during the interview.\n\nDo not invent test scenarios that were never requested or clarified merely to make the Specification look complete.\n\n---\n\n# Existing tickets\n\nWhen updating an existing ticket:\n\n- preserve useful existing content,\n- preserve Specification sections that are not being changed,\n- do not create duplicate Specification blocks,\n- do not overwrite unrelated description content.\n\nIf asked to create a ticket whose title closely matches an existing ticket, tell the user about the possible duplicate and offer to update the existing ticket instead.\n\n---\n\n# Output structure\n\nThe resulting Specification represents:\n\nSpecification\n\nProblem statement\n\u003Ccontent>\n\nSolutioning\n\u003Ccontent>\n\nTest cases\n\u003Ccontent>\n\nImplementation details\n\u003Ccontent, only when applicable>\n\nOut of scope\n\u003Ccontent, only when applicable>\n\nUse these heading names exactly because downstream PR specification review depends on a stable Specification contract.\n\n---\n\n# Writing to Xyne Spaces\n\nWhen the Xyne Spaces ticket tools support a structured `specification` parameter, pass the Specification sections as structured values.\n\nFor example, conceptually:\n\nspecification:\n- problemStatement\n- solutioning\n- testCases\n- implementationDetails\n- outOfScope\n\nLet the Spaces backend render and merge the Specification into the stored rich-text ticket description.\n\nDo not perform a read-modify-write of the complete ticket description merely to add or update the Specification.\n\nTicket descriptions may contain rich-text HTML.\n\nIf the description returned to the agent has been converted to plain text, reading it, appending the Specification, and writing the complete description back can destroy:\n\n- formatting,\n- lists,\n- links,\n- images,\n- attachments,\n- and other rich-text content.\n\nThe Spaces backend should remain responsible for merging structured Specification content into the raw stored ticket description.\n\n---\n\n# Core rules\n\nAlways follow these rules:\n\n1. Understand the ticket context before interviewing the user.\n2. Questions must vary based on the ticket context.\n3. Do not mechanically ask a fixed questionnaire.\n4. Ask only questions that materially improve the Specification.\n5. Ask the initial questions in one batch.\n6. Wait for the user before generating the final Specification.\n7. Never invent requirement information.\n8. Do not derive requirement intent solely from implementation.\n9. Required sections must be meaningfully populated.\n10. Optional sections may be omitted.\n11. Show the Specification draft before writing it.\n12. Obtain user confirmation before updating the ticket.\n13. Preserve unrelated existing ticket content.\n14. Keep the Specification headings stable for downstream PR review.\n","# Xyne Spaces Ticket Specification\n\nUse this skill when creating, generating, regenerating, or updating a Xyne Spaces ticket specification.\n\nThe purpose of the Specification is to capture the requirement intent so that:\n\n- a developer can understand what is expected,\n- a PR reviewer can understand what was requested,\n- and the implementation can later be reviewed against that requirement.\n\nThe Specification must come from requirement context and user clarification.\n\nDo not generate requirement intent merely by describing the implementation.\n\n---\n\n## Specification sections\n\nUse these headings exactly.\n\n### Required\n\n#### Problem statement\n\nDescribe what is wrong, missing, or needed and who or what is affected.\n\nThe Problem statement should explain why this work exists.\n\n#### Solutioning\n\nDescribe the expected functional behaviour after the change is shipped.\n\nFocus on what must be true functionally, not how the developer should implement it.\n\n#### Test cases\n\nDescribe the observable scenarios that confirm the requirement works.\n\nTest cases should make it possible for a developer or reviewer to understand what behaviours need to succeed.\n\n### Optional\n\n#### Implementation details\n\nTechnical details, constraints, or implementation context that are useful to the developer or reviewer.\n\nDo not require this section when there is nothing meaningful to record.\n\n#### Out of scope\n\nAnything explicitly excluded from this ticket.\n\nDo not invent non-goals simply to populate this section.\n\n---\n\n# Before the interview\n\nBefore creating or regenerating a Specification, first understand the requirement context.\n\nIf this is an existing ticket:\n\n1. Fetch the ticket using `spaces-tickets`.\n2. Read:\n   - ticket title,\n   - ticket type, when available,\n   - existing description,\n   - existing Specification, when present,\n   - other relevant requirement context available with the ticket.\n3. Use this information to understand what the ticket is about.\n4. Use it to formulate better and more relevant clarification questions.\n5. Do NOT create or update the Specification yet.\n\nThe existing ticket description is context.\n\nIt does NOT automatically replace the requirement interview.\n\nDo not silently convert an existing one-line description into a complete Specification.\n\nDo not treat the implementation, PR description, commit messages, or code diff as the source of truth for requirement intent.\n\n---\n\n# Mandatory contextual interview\n\nBefore creating or regenerating a Specification, interview the user.\n\nThe Specification sections are fixed:\n\nREQUIRED\n- Problem statement\n- Solutioning\n- Test cases\n\nOPTIONAL\n- Implementation details\n- Out of scope\n\nHowever, the interview questions are NOT fixed.\n\nThe questions MUST be adapted to the context of the specific ticket.\n\nDo not mechanically ask the same questions for every ticket.\n\n---\n\n## Generate questions from the ticket context\n\nUse the ticket title, ticket type, description, current conversation, and other available requirement context to determine what needs clarification.\n\nGenerate questions that help you confidently populate the Specification.\n\nPrioritise questions that clarify:\n\n- what was originally requested,\n- what problem needs to be solved,\n- who or what is affected,\n- expected user-visible behaviour,\n- important product decisions,\n- edge cases,\n- failure behaviour,\n- success criteria,\n- expected test scenarios,\n- technical constraints when relevant,\n- explicit non-goals or boundaries.\n\nDo NOT simply turn the section headings into generic questions.\n\nBad:\n\n- What is the problem statement?\n- What is the solutioning?\n- What are the test cases?\n- What are the implementation details?\n- What is out of scope?\n\nInstead, understand the requirement first and ask concrete questions about the specific feature, bug, or task.\n\n---\n\n## Example — bug\n\nTicket:\n\n> Mention picker shows no users in sandbox.\n\nUseful contextual questions might include:\n\n- When the normal user lookup fails or returns no users, what should the picker show instead?\n- What should the user see if the fallback also fails?\n- Under which environments or conditions is this expected to work?\n- Which scenarios should we verify before considering the bug fixed?\n- Is anything related to mention behaviour explicitly outside this fix?\n\nThese are examples only.\n\nDo not reuse these questions when they are irrelevant to the actual ticket.\n\n---\n\n## Example — feature\n\nTicket:\n\n> Add people to an existing conversation.\n\nUseful contextual questions might include:\n\n- Who should be allowed to add another person to a conversation?\n- What conversation history should a newly added person be able to access?\n- How should files and attachments behave for the newly added person?\n- What should happen if adding the participant fails?\n- What scenarios must work for this feature to be considered complete?\n- Is anything explicitly outside the scope of this feature?\n\nThese are examples only.\n\nGenerate questions appropriate to the actual feature.\n\n---\n\n## Example — technical task\n\nTicket:\n\n> Move ticket specification merging to the Spaces backend.\n\nUseful contextual questions might include:\n\n- What problem with the current specification update flow are we trying to eliminate?\n- What should happen when the ticket already contains a Specification?\n- What existing ticket-description content must be preserved?\n- What should happen if the Specification is malformed or duplicated?\n- What cases should be tested to prove that existing rich-text content is preserved?\n- Are any creation or update paths explicitly outside this change?\n\nThese are examples only.\n\n---\n\n# Question selection\n\nAsk only questions that materially improve the Specification.\n\nDo not ask the user to repeat information they have already explicitly provided.\n\nFor example, if the user has already said:\n\n> Newly added users should see previous messages but should not get access to old attachments.\n\ndo not ask:\n\n> Should newly added users see previous messages?\n\nThat decision is already established.\n\nInstead, identify what remains unclear.\n\nFor example:\n\n- Who is allowed to add people?\n- What happens if adding someone fails?\n- Does the same behaviour apply to private conversations?\n- What scenarios must be verified?\n\n---\n\n## Number of questions\n\nThe number of questions should vary based on the complexity and ambiguity of the ticket.\n\nA simple ticket may require only 2–3 clarification questions.\n\nA complex feature may require 6–10 questions.\n\nDo not ask additional questions merely to reach a fixed number.\n\nAsk the minimum useful set of questions needed to establish the required Specification.\n\n---\n\n## Ask questions in one batch\n\nAsk the initial clarification questions together.\n\nDo not unnecessarily ask one question, wait for an answer, then ask the next question.\n\nThe user should be able to answer the main interview in one response.\n\nAfter asking the questions:\n\nSTOP.\n\nWait for the user's response.\n\nDo NOT create or update the Specification in the same turn in which you ask the interview questions.\n\nDo NOT call `spaces-update-ticket` yet.\n\n---\n\n# Requirement source\n\nProblem statement, Solutioning, and Test cases must represent what was requested or explicitly clarified by the user.\n\nDo not construct the requirement by simply summarising:\n\n- the implementation,\n- PR diff,\n- changed files,\n- commit messages,\n- PR description,\n- or completed code.\n\nGood:\n\n> When the normal user lookup fails, what should the picker show instead?\n\nBad:\n\n> I see that you implemented a fallback to channel participants. Should the requirement say that?\n\nThe implementation may be used as additional context for asking sharper questions.\n\nIt must not become the source of truth for requirement intent.\n\nA Specification derived only from the implementation will naturally resemble the implementation and therefore cannot reliably reveal something that was requested but never implemented.\n\n---\n\n# After the user answers\n\nAfter receiving the user's answers:\n\n1. Map the answers into:\n   - Problem statement,\n   - Solutioning,\n   - Test cases,\n   - Implementation details, when applicable,\n   - Out of scope, when applicable.\n\n2. Use the user's answers and explicitly provided requirement context as the source of truth.\n\n3. Check whether all required sections can now be populated meaningfully.\n\nRequired:\n- Problem statement\n- Solutioning\n- Test cases\n\n4. Do not invent missing information.\n\n5. If required information remains missing or ambiguous, ask only the necessary follow-up questions.\n\n6. Do not repeat the entire interview.\n\n7. Optional sections may be omitted when they are not applicable or the user has provided no meaningful information for them.\n\n---\n\n# Missing information\n\nNever invent a plausible answer.\n\nFor required sections:\n\n- If the user's answer is unclear, ask for clarification.\n- If an important requirement decision remains unresolved, ask about it.\n- Do not consider the Specification complete until the required information is sufficiently clear.\n\nFor optional sections:\n\n- Omit the section if there is no meaningful information.\n- Do not generate filler merely so every heading appears.\n\nDo not use \"Not specified\" to bypass missing required information when the user is available to clarify it.\n\n---\n\n# Draft before updating\n\nOnce sufficient information has been collected, draft the complete Specification.\n\nShow the draft to the user before modifying the ticket.\n\nUse:\n\n## Specification\n\n### Problem statement\n\u003Cdraft>\n\n### Solutioning\n\u003Cdraft>\n\n### Test cases\n\u003Cdraft>\n\n### Implementation details\n\u003Cdraft, only when applicable>\n\n### Out of scope\n\u003Cdraft, only when applicable>\n\nThen ask the user to confirm or correct the Specification.\n\nDo NOT update the ticket until the user confirms the draft.\n\n---\n\n# Confirmation\n\nThe user must have an opportunity to review the generated Specification before it is written to the ticket.\n\nIf the user requests changes:\n\n1. Update the draft.\n2. Show the changed Specification when necessary.\n3. Obtain confirmation.\n\nAfter the user confirms, create or update the ticket Specification.\n\n---\n\n# Existing Specification edits\n\nA full interview is required when:\n\n- creating the first Specification,\n- regenerating the Specification,\n- substantially rewriting the Specification,\n- or the existing Specification does not contain enough requirement information.\n\nA full interview is NOT required for an explicit targeted edit.\n\nExample:\n\n> Add \"notifications are not part of this work\" to Out of scope.\n\nIn that case:\n\n1. Fetch the existing ticket.\n2. Preserve the existing Specification.\n3. Apply only the requested change.\n4. Show the resulting change\u002Fspecification to the user when appropriate.\n5. Update after confirmation.\n\nDo not ask the user to answer Problem statement, Solutioning, and Test cases again for a simple targeted edit.\n\nIf the user says:\n\n- redo the spec,\n- regenerate the spec,\n- rewrite the spec,\n- create a new spec,\n\nrun the contextual interview again.\n\n---\n\n# Pasted requirement or external-agent content\n\nIf the user pastes:\n\n- a requirement,\n- PRD,\n- product discussion,\n- Slack\u002Fthread conversation,\n- notes,\n- output from another coding agent,\n- or other requirement material,\n\nuse it as context.\n\n1. Extract information that is explicitly established.\n2. Map what is already known to the Specification sections.\n3. Identify important ambiguities, missing decisions, edge cases, and missing required information.\n4. Generate contextual clarification questions for those gaps.\n5. Do not ask the user to repeat information already explicitly provided.\n6. Do not silently convert implementation-derived observations into requirement intent.\n7. Wait for the user's answers.\n8. Map the confirmed information into the Specification.\n9. Follow the normal draft → confirmation → update flow.\n\n---\n\n# Writing style\n\nWrite Specifications for developers and PR reviewers.\n\nBe concise.\n\nBe concrete.\n\nPrefer observable behaviour.\n\nAvoid vague requirements.\n\nBad:\n\n> Improve the mention picker.\n\nGood:\n\n> When the primary user lookup returns no users, the mention picker should fall back to eligible channel participants.\n\nAvoid unnecessary implementation prescriptions in Problem statement and Solutioning.\n\nProblem statement should explain the problem.\n\nSolutioning should explain the expected functional outcome.\n\nImplementation details should contain technical implementation constraints when those details actually matter.\n\n---\n\n# Test cases\n\nTest cases should describe meaningful scenarios.\n\nBad:\n\n> Test the feature.\n\nBetter:\n\n> When the primary user lookup returns users, the picker displays those users normally.\n\n> When the primary lookup returns no users, the picker displays eligible channel participants.\n\n> When both the primary lookup and fallback fail, the user sees the specified error state.\n\nInclude positive, failure, or edge scenarios when they were established during the interview.\n\nDo not invent test scenarios that were never requested or clarified merely to make the Specification look complete.\n\n---\n\n# Existing tickets\n\nWhen updating an existing ticket:\n\n- preserve useful existing content,\n- preserve Specification sections that are not being changed,\n- do not create duplicate Specification blocks,\n- do not overwrite unrelated description content.\n\nIf asked to create a ticket whose title closely matches an existing ticket, tell the user about the possible duplicate and offer to update the existing ticket instead.\n\n---\n\n# Output structure\n\nThe resulting Specification represents:\n\nSpecification\n\nProblem statement\n\u003Ccontent>\n\nSolutioning\n\u003Ccontent>\n\nTest cases\n\u003Ccontent>\n\nImplementation details\n\u003Ccontent, only when applicable>\n\nOut of scope\n\u003Ccontent, only when applicable>\n\nUse these heading names exactly because downstream PR specification review depends on a stable Specification contract.\n\n---\n\n# Writing to Xyne Spaces\n\nWhen the Xyne Spaces ticket tools support a structured `specification` parameter, pass the Specification sections as structured values.\n\nFor example, conceptually:\n\nspecification:\n- problemStatement\n- solutioning\n- testCases\n- implementationDetails\n- outOfScope\n\nLet the Spaces backend render and merge the Specification into the stored rich-text ticket description.\n\nDo not perform a read-modify-write of the complete ticket description merely to add or update the Specification.\n\nTicket descriptions may contain rich-text HTML.\n\nIf the description returned to the agent has been converted to plain text, reading it, appending the Specification, and writing the complete description back can destroy:\n\n- formatting,\n- lists,\n- links,\n- images,\n- attachments,\n- and other rich-text content.\n\nThe Spaces backend should remain responsible for merging structured Specification content into the raw stored ticket description.\n\n---\n\n# Core rules\n\nAlways follow these rules:\n\n1. Understand the ticket context before interviewing the user.\n2. Questions must vary based on the ticket context.\n3. Do not mechanically ask a fixed questionnaire.\n4. Ask only questions that materially improve the Specification.\n5. Ask the initial questions in one batch.\n6. Wait for the user before generating the final Specification.\n7. Never invent requirement information.\n8. Do not derive requirement intent solely from implementation.\n9. Required sections must be meaningfully populated.\n10. Optional sections may be omitted.\n11. Show the Specification draft before writing it.\n12. Obtain user confirmation before updating the ticket.\n13. Preserve unrelated existing ticket content.\n14. Keep the Specification headings stable for downstream PR review.\n",{"name":5,"description":7},"main","fbefa477cb2b164d49af3ffe1b51c219b3cae6d1","apps\u002Fxyne-claw\u002Fspec-skills\u002Fspec-task",573,"https:\u002F\u002Fgithub.com\u002Fjuspay\u002Fxyne-spaces","https:\u002F\u002Fxyne.ai","https:\u002F\u002Fgithub.com\u002Fjuspay","https:\u002F\u002Fwww.juspay.in"," Design to simplify. Revolutionizing digital payments.","2026-11-15 05:06:40",{"source":8,"skillName":5,"commands":35,"prompt":45,"zipAvailable":46},[36,39,42],{"tool":37,"command":38},"npx","npx skills add juspay\u002Fxyne-spaces --skill \"Ticket Specs\"",{"tool":40,"command":41},"bunx","bunx skills add juspay\u002Fxyne-spaces --skill \"Ticket Specs\"",{"tool":43,"command":44},"pnpm","pnpm dlx skills add juspay\u002Fxyne-spaces --skill \"Ticket Specs\"","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fjuspay\u002Fxyne-spaces\u002Fblob\u002Fmain\u002Fapps\u002Fxyne-claw\u002Fspec-skills\u002Fspec-task\u002FSKILL.md",true,[],[49,56,63,70,77,84],{"skillNo":50,"name":51,"nameZh":6,"description":52,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":53,"categories":54,"keywords":55,"publishedAt":19},"01M7A02Z2C9ECN1C13PE67A3DW","whoami","Show current Xyne Claw auth identity.","https:\u002F\u002Fgithub.com\u002Fjuspay\u002Fxyne-spaces\u002Fblob\u002Fmain\u002Fpackages\u002Fxyne-claw-mcp\u002Fskills\u002Fwhoami\u002FSKILL.md",[],[],{"skillNo":57,"name":58,"nameZh":6,"description":59,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":60,"categories":61,"keywords":62,"publishedAt":19},"01M7A02XW56HSFGZ9PNFY3AWDT","sessions","List Xyne Claw sessions.","https:\u002F\u002Fgithub.com\u002Fjuspay\u002Fxyne-spaces\u002Fblob\u002Fmain\u002Fpackages\u002Fxyne-claw-mcp\u002Fskills\u002Fsessions\u002FSKILL.md",[],[],{"skillNo":64,"name":65,"nameZh":6,"description":66,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":67,"categories":68,"keywords":69,"publishedAt":19},"01M7A02WQ3EFMNYBS0QF61JFJ6","run","Run a Xyne Claw agent.","https:\u002F\u002Fgithub.com\u002Fjuspay\u002Fxyne-spaces\u002Fblob\u002Fmain\u002Fpackages\u002Fxyne-claw-mcp\u002Fskills\u002Frun\u002FSKILL.md",[],[],{"skillNo":71,"name":72,"nameZh":6,"description":73,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":74,"categories":75,"keywords":76,"publishedAt":19},"01M7A02VC1F69GS89JJBC9HCT5","login","Log the user into Xyne Claw.","https:\u002F\u002Fgithub.com\u002Fjuspay\u002Fxyne-spaces\u002Fblob\u002Fmain\u002Fpackages\u002Fxyne-claw-mcp\u002Fskills\u002Flogin\u002FSKILL.md",[],[],{"skillNo":78,"name":79,"nameZh":6,"description":80,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":81,"categories":82,"keywords":83,"publishedAt":19},"01M7A02SHD94QXZ4D37P9EMW9V","agents","List available Xyne Claw agents.","https:\u002F\u002Fgithub.com\u002Fjuspay\u002Fxyne-spaces\u002Fblob\u002Fmain\u002Fpackages\u002Fxyne-claw-mcp\u002Fskills\u002Fagents\u002FSKILL.md",[],[],{"skillNo":85,"name":86,"nameZh":6,"description":87,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":88,"categories":89,"keywords":90,"publishedAt":19},"01M7A02KCZS8QEARG526JJAQBZ","explanation-document","Structure a codebase\u002Fschema explanation as a self-contained HTML document ordered for a READER — mental model and flows first, deep per-component reference last — not in the order you explored it.","https:\u002F\u002Fgithub.com\u002Fjuspay\u002Fxyne-spaces\u002Fblob\u002Fmain\u002Fapps\u002Fxyne-claw\u002Funderstanding-skills\u002Fexplanation-document\u002FSKILL.md",[],[]]