capture-decision-context
Distill the current conversation into a durable, implementation-light decision note or compact GitHub issue draft that preserves motivation, evidence…
Skill by corymhall
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/corymhall-capture-decision-context
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: productivity
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/capture-decision-context
What's inside
- Skill name
capture-decision-context- Description
- Distill the current conversation into a durable, implementation-light decision note or compact GitHub issue draft that preserves motivation, evidence, rationale, tradeoffs, accepted limitations, and open questions without turning session details into an authoritative specification or implementation plan.
- License
- MIT
- Instructions
- 85 lines in SKILL.md · view on GitHub
Files (2)
- SKILL.md
- openai.yaml · agents/
SKILL.md (first lines)
# Capture Decision Context Preserve the parts of the current conversation that a future agent or contributor would otherwise lose. Produce a decision/context artifact, not a complete implementation specification. Use the conversation as the source of truth. Consult the repository only to confirm named facts, links, current status, or the appropriate destination; do not restart planning or broaden the investigation. If a repository-note destination is unclear, ask the user where it should live. Capture only what adds durable value: - the problem, motivation, and desired observable outcome - evidence that materially shaped the discussion - accepted decisions and why they were chosen - rejected alternatives when the reason matters later - constraints, non-goals, tradeoffs, accepted limitations, and safe fallbacks - genuinely open questions - concise status, validation evidence, and high-value references when useful Distinguish user-accepted decisions from assistant proposals, rejected ideas, and unresolved questions. Do not promote suggestions or speculation into requirements. Keep the artifact implementation-light. Avoid file-by-file tasks, exhaustive edge-case or test matrices, detailed pseudocode, speculative requirements, and facts that are easy to rediscover. Keep an implementation detail only when it records an external contract, settled constraint, consequential tradeoff, or rationale for rejecting an alternative. ## Decision Note Adapt this shape and omit empty sections: ```markdown # <Topic> Decision Context ## Motivation and Desired Outcome ## Relevant Evidence ## Decisions and Rationale ## Tradeoffs, Limitations, and Non-Goals ## Open Questions ## Current Status ## References ``` ## GitHub Issue Draft When the user wants to snapshot the investigation as an issue, keep the visible body compact and move supporting context into an expandable section. Include a proposed direction only when the user accepted it or label it explicitly as provisional. Include `Done when` only when it clarifies the observable outcome. ```markdown ## Problem ## Proposed direction ## Done when <details> <summary>Context and evidence</summary> ### Relevant evidence and history ### Tradeoffs, limitations, and rejected directions </details> ``` Omit inapplicable headings. Make every visible sentence change the reader's understanding of the current problem or desired outcome. Keep only the most relevant history, evidence, rationale, tradeoffs, limitations, rejected directions, and links in the details section; do not copy the full session note. Label API or architecture sketches as proposals unless the user settled them. Draft in chat by default. Create or post the issue only when the user explicitly asks, and do not also create a repository context note unless requested. Prefer a short concrete artifact over comprehensive session summarization. Preserve uncertainty rather than resolving it by inference. A future agent should use the artifact with the current repository and maintainer conversation, not treat it as authoritative implementation instructions.
From corymhall/toolkit, MIT.
More from corymhall/toolkit
- 0.529score
harness-evaluation
corymhall
Evaluate real Pi and Codex sessions against centralized criteria for personal and repo-local skills, identify whether skills helped, were neutral, or made…
SkillBFreshness: BClaude Code +9 - 0.526score
triage-provider-issue
corymhall
Triage Pulumi provider issues from evidence to disposition or next action.
SkillBFreshness: BClaude Code +9 - 0.526score
skill-authoring
corymhall
Create, revise, de-slopify, or review Agent Skills while preserving useful behavior and requiring evidence for added prompt weight.
SkillBFreshness: BClaude Code +9 - 0.526score
go-development
corymhall
Implement and refactor Go where package design, abstractions, ownership, errors, concurrency, or Go-specific test seams require manual judgment.
SkillBFreshness: BClaude Code +9 - 0.524score
commit
corymhall
Create a well-formed git commit from current changes using session history for rationale and summary; use only when the user explicitly asks to create or…
SkillBFreshness: BClaude Code +9 - 0.521score
collaborate
corymhall
Establish a pair-programming working agreement for the session.
SkillBFreshness: BClaude Code +9 - 0.519score
aws-documentation-mcp
corymhall
Search, read, and cite official AWS documentation through the AWS Labs Documentation connection in executor.
SkillBFreshness: BClaude Code +9 - 0.479score
stage-terraform-provider-repro
corymhall
Stage durable Terraform-side repros for bridged-provider issues.
SkillCFreshness: CClaude Code +9 - 0.479score
stage-pulumi-provider-repro
corymhall
Stage durable Pulumi-side repros for provider issues.
SkillCFreshness: CClaude Code +9 - 0.479score
pulumi-upgrade-provider
corymhall
Automate Pulumi provider repo upgrades with the `upgrade-provider` tool.
SkillCFreshness: CClaude Code +9 - 0.479score
bridge-parity-investigation
corymhall
Investigate established Pulumi-vs-Terraform parity gaps with bridge cross-tests.
SkillCFreshness: CClaude Code +9 - 0.476score
neovim-plugin-development
corymhall
Implement, refactor, modernize, and debug Neovim plugins in Lua using official runtimepath/module-loading, startup/lazy-loading, keymap/autocmd, health…
SkillCFreshness: CClaude Code +9
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Aug 3, 2026 (2 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/corymhall/toolkit
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from corymhall
- 0.529score
harness-evaluation
corymhall
Evaluate real Pi and Codex sessions against centralized criteria for personal and repo-local skills, identify whether skills helped, were neutral, or made…
SkillBFreshness: BClaude Code +9 - 0.526score
go-development
corymhall
Implement and refactor Go where package design, abstractions, ownership, errors, concurrency, or Go-specific test seams require manual judgment.
SkillBFreshness: BClaude Code +9 - 0.526score
triage-provider-issue
corymhall
Triage Pulumi provider issues from evidence to disposition or next action.
SkillBFreshness: BClaude Code +9 - 0.526score
skill-authoring
corymhall
Create, revise, de-slopify, or review Agent Skills while preserving useful behavior and requiring evidence for added prompt weight.
SkillBFreshness: BClaude Code +9 - 0.524score
commit
corymhall
Create a well-formed git commit from current changes using session history for rationale and summary; use only when the user explicitly asks to create or…
SkillBFreshness: BClaude Code +9 - 0.521score
collaborate
corymhall
Establish a pair-programming working agreement for the session.
SkillBFreshness: BClaude Code +9
Cite this page
capture-decision-context skill. Agent Mart. https://agentmart.dev/skills/corymhall-capture-decision-context (last updated 2026-10-24)