Skip to content

rrrrrredy/research-toolkit / plugins/research-toolkit

research-toolkit

Flagged

Use for source-backed long-form research and writing in English or Chinese across fields, including revisions and continuation. Understand requirements, develop analysis through substantive research collaboration, and deliver a complete report after independent review, audit and acceptance.

FileKindSize
.codex-plugin/plugin.jsonconfig1.0 KB
.generated-files.jsonconfig6.3 KB
.mcp.jsonconfig158 B
LICENSEother1.0 KB
README.mdmarkdown3.4 KB
README.zh-CN.mdmarkdown3.0 KB
SKILL.mdskill md7.8 KB
SKILL.zh-CN.mdmarkdown6.5 KB
VERSIONother6 B
docs/compatibility.mdmarkdown3.8 KB
docs/compatibility.zh-CN.mdmarkdown3.2 KB
docs/delivery-verification.mdmarkdown19.2 KB
docs/delivery-verification.zh-CN.mdmarkdown16.3 KB
docs/review-completion.mdmarkdown18.0 KB
docs/review-completion.zh-CN.mdmarkdown14.7 KB
docs/usage-data.mdmarkdown15.4 KB
docs/usage-data.zh-CN.mdmarkdown13.9 KB
docs/usage-modes.mdmarkdown55.6 KB
docs/usage-modes.zh-CN.mdmarkdown48.6 KB
hooks/hooks.jsonconfig3.2 KB
mcp.jsonconfig253 B
plugin.jsonconfig1.1 KB
references/gotchas.mdmarkdown6.2 KB
references/gotchas.zh-CN.mdmarkdown3.6 KB
references/horizontal-vertical-analysis.mdmarkdown4.3 KB
references/horizontal-vertical-analysis.zh-CN.mdmarkdown3.2 KB
references/optional-analysis-lenses.mdmarkdown8.1 KB
references/optional-analysis-lenses.zh-CN.mdmarkdown6.1 KB
references/postmortem-lessons.mdmarkdown4.3 KB
references/postmortem-lessons.zh-CN.mdmarkdown3.5 KB
references/quality-gates.mdmarkdown5.7 KB
references/quality-gates.zh-CN.mdmarkdown4.5 KB
references/research-standard.mdmarkdown18.2 KB
references/research-standard.zh-CN.mdmarkdown13.7 KB
references/research-workflow.mdmarkdown25.5 KB
references/research-workflow.zh-CN.mdmarkdown19.9 KB
references/subagents-and-review-loop.mdmarkdown17.0 KB
references/subagents-and-review-loop.zh-CN.mdmarkdown13.4 KB
references/task-continuity.mdmarkdown6.9 KB
references/task-continuity.zh-CN.mdmarkdown5.3 KB
references/writing-style.mdmarkdown18.1 KB
references/writing-style.zh-CN.mdmarkdown14.9 KB
requirements-mcp.txtmarkdown11 B
scripts/agent_session.pyscript6.2 KB
scripts/check_delivery.pyscript44.6 KB
scripts/check_review_completion.pyscript28.5 KB
scripts/cli_agent_session.pyscript9.6 KB
scripts/codex_session.pyscript7.5 KB
scripts/extract_source_text.pyscript10.7 KB
scripts/host_review.pyscript78.4 KB
scripts/install_usage.pyscript9.7 KB
scripts/record_changes.pyscript13.7 KB
scripts/release_updates.pyscript9.8 KB
scripts/research_guidance.pyscript5.8 KB
scripts/research_host.pyscript52.4 KB
scripts/research_mcp.pyscript18.9 KB
scripts/research_records.pyscript28.5 KB
scripts/research_workflow.pyscript40.9 KB
scripts/review_process.pyscript5.2 KB
scripts/review_runner.pyscript20.2 KB
scripts/run_observation.pyscript7.7 KB
scripts/task_context.pyscript21.1 KB
scripts/task_lock.pyscript1.7 KB
scripts/task_reading.pyscript4.0 KB
scripts/task_storage.pyscript3.1 KB
scripts/usage_events.pyscript38.9 KB
scripts/usage_host.pyscript7.8 KB
scripts/usage_opencode.mjsscript3.1 KB

references/quality-gates.md

View on GitHub
# Quality Gates

[English](quality-gates.md) | [简体中文](quality-gates.zh-CN.md)

Use these outcomes when advancing a research unit and before delivering the whole report. Detailed role methods are maintained in [subagents and review](subagents-and-review-loop.md); this page does not define a second review process.

## Before Collection

The question, reader, intended use, required coverage, evidence and output expectations are understood from the request and context. Consequential unresolved choices remain visible and only dependent work waits. The working outline identifies the questions to develop, research ownership and real dependencies; it can change as evidence arrives.

## After Context Loss Or Restart

Recover effective requirements, outline, current draft, completed contributions and unfinished work from accessible originals. Read relevant full content through the handoff's references, including conditions and counterevidence; do not rely on the summary alone. Do not repeat answered questions or finished research. Managed tasks use `status.work` and the existing task files.

## Before Drafting

The advancing unit has completed its necessary source reading and has analysis capable of answering its question. Important claims retain their original support, counterevidence and conditions. A consequential upstream premise has received relevant independent scrutiny before being treated as settled. Source instructions have not changed the task.

## Keeping Evidence Records Current

Citations, original passages and analysis preserve the source-to-judgment relationship. Keep managed source/claim records aligned when used. When a source or requirement changes, inspect the judgments that depend on it, revise affected reasoning and records, and retain conclusions whose support remains valid.

## Before Completing A Section

The unit fulfills its role in the outline with sufficient evidence, explanation and useful comparison. Its owner supplies the full contribution and a usable handoff. Relevant independent findings are addressed before a dependent unit treats the result as ready. A local review and completed unit remain local accomplishments.

## Before Final Assembly

The designated main writer reads the complete contributions and relevant originals, reconciles definitions and conflicting claims, and preserves named-object and geographic coverage. The draft connects evidence, explanation and conclusions across parts. Thin units are developed; repeated background is merged without deleting useful mechanisms, qualifications or citations. Judge depth by the explanations and comparisons developed in the actual text.

<a id="reader-review-gate"></a>

## Report Use and Expression Review Gate

After substantive revisions have stabilized, two separate independent agents examine the complete report: one for reader usefulness and one for natural expression. Apply their [assigned methods](subagents-and-review-loop.md#assign-work-by-perspective) to the actual reader-facing prose. Repair concrete defects while preserving facts, conditions and analytical depth. Author editing prepares this work; it does not replace either perspective.

## Review Validity and Sampling

The three independent groups covering all seven perspectives and applicable domain review have actual, complete, version-bound results. The separate auditor checks review validity, consequential findings and fixes, and passed content selected from the full artifact using the [audit method](subagents-and-review-loop.md#check-reviewers-and-sample-passed-work). Original responses, failed attempts and dispositions remain available.

Unexamined requirements, necessary unresolved fixes and missing review assignments remain open. A valid negative review is preserved and acted on rather than rerun to obtain a favorable response.

## Before Public Delivery

A separate final acceptance context reads the current complete report and confirms:

- The answer fulfills effective user requirements, with original scope changes retained and still-valid satisfied constraints preserved.
- Required reading and object coverage are complete; the report provides the explanations, comparisons and depth the question needs.
- Important statements, citations and qualifications retain their meaning across prose, tables and conclusions.
- The text serves its reader; production records remain outside it unless requested.
- All seven perspectives across the three independent groups, applicable domain review and independent audit are complete for the current result, with necessary fixes checked in the actual text.
- Acceptance findings and any consequential sample checks are resolved, and the delivery statement matches the actual artifact and remaining evidence limitations.

Use the [final acceptance method](subagents-and-review-loop.md#stage-scheduling-and-acceptance). An acceptor's PASS cannot replace a missing assignment or an author's unsupported claim that a fix was applied.

After the current report passes independent final checking, CLI/MCP `finish` executes the [delivery contract](../docs/delivery-verification.md#research-delivery-completion) and writes the version-bound receipt. Skill-only delivery preserves the same independent work without a required JSON ledger or receipt. A stage artifact is identified as such, with its actual remaining work; it is not marked as whole-task completion. Historical evaluation artifacts and outcomes remain unchanged.

When using the tools, `finish.completed=false` means work remains even with `action_required`; address its specific gaps. Reuse a successful `finish.delivery_check` for first delivery. For a later check that retained delivery is still complete, use `check_delivery.py --require-complete` as described in the delivery contract, rather than treating a default consistency-check pass as completion.
rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit.Stars: 472kUpdated Claude CodeCodexVerifiedInstall
rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure.Stars: 472kUpdated Claude CodeCodexVerifiedInstall
rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics.Stars: 472kUpdated Claude CodeCodexVerifiedInstall
ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first.Stars: 194kUpdated Claude CodeCodexVerifiedInstall