원클릭으로 Manus에서 모든 스킬 실행

html-experiment-report

스타2

포크0

업데이트2026년 6월 29일 06:29

Use when producing a browser-readable HTML experiment or Eval report; first decide the primary figure, then plan and run an evidence-backed report renderer while keeping domain authority in the original tool.

설치

Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.

Manus에서 실행

출처

iwashita-nozomu

iwashita-nozomu/agent-canon

GitHub 저장소 열기 Creator 저장소 보기

다운로드

Manus에서 실행

HTML Experiment Report

Tool Commands

Use the command packet before applying this skill's workflow:

python3 tools/agent_tools/skill_tool_commands.py show --skill html-experiment-report --format text

Execute the required and task-matching conditional commands that the packet prints.

Read agents/skills/html-experiment-report.md.
Use $structure-planning before renderer or experiment implementation to fix the first figure, source-to-structure map, metric contract, section order, and invalid interpretations.
Use $html-output for polished HTML layout, optional $imagegen assets, existing preview-server reuse, python3 -m http.server --bind 0.0.0.0 publication, curl -fsS validation, and local/external browser URLs.
Survey existing assets first: relevant skills, tool catalog entries, workflow docs, report helpers, previous run artifacts, and existing experiment scripts.
Write a responsibility analysis before implementation: raw evidence owner, renderer owner, domain decision owner, generated artifact path, and the invalid authority drift to avoid.
Name the first figure through the structure contract before running the experiment. Record its question, required data, metric denominator, directionality, and invalid interpretations.
Derive the experiment plan from that figure: evidence-producing command, renderer command, HTML output path, blocked-provider behavior, and validation gates.
Reuse existing producers and helpers. Add a report-specific renderer or adapter only when the HTML artifact needs one.
Keep generated SQLite, JSON, and HTML artifacts under ignored run paths such as reports/agents/<run-id>/; do not make generated HTML a policy truth.
Put the primary figure first in the HTML, followed by source packet, observations, interpretation, limitations, provenance, and next action.
For semantic-index provider comparison, use the figure Provider Delta To Shared Candidate Logic and state that LLM latent vectors may change retrieval/ranking deltas but do not create labels, ownership, or merge authority.
Validate with targeted renderer tests, docs checks for changed Markdown, and catalog/dependency checks for changed tool or skill wiring.
Record closeout tokens: html_experiment_report=complete, html_primary_figure=<figure-name>, structure_contract=<path-or-inline>, html_source_artifact=<path>, html_report_artifact=<path>, html_domain_authority=<tool-or-doc>, and html_output=complete, html_server_mode=<reuse|started|not_requested|blocked>, and html_invalid_interpretations_recorded=yes.

이 저장소의 다른 Skills

같은 저장소

agent-learning

iwashita-nozomu/agent-canon

Use when agent-side working philosophy, interaction lessons, task retrospectives, repeated routing misses, missed skill invocation, or recurrence-prevention feedback should be logged without mixing them into user preferences.

2026-07-082

algorithm-proof-exploration

iwashita-nozomu/agent-canon

Use when exploring, refactoring, or choosing an algorithm under proof obligations; builds JIT-canonical IR, lemma dependency graphs, algorithmic blocker frontiers, and algorithm-change guidance before handing terminal proof work to formal-proof-workflow.

2026-07-082

computational-optimization

iwashita-nozomu/agent-canon

Use when designing, implementing, reviewing, or diagnosing numerical optimization, solvers, preconditioners, convergence, gradients, Jacobians, Hessians, KKT conditions, tolerances, or optimization benchmarks; fixes the mathematical and validation contract before code or experiment changes.

2026-07-082

test-design

iwashita-nozomu/agent-canon

Use when a change needs oracle/spec-risk classification or resilient, adversarial static test design, including behavior contracts, oracle choice, property/metamorphic candidates, mutation adequacy, or brittle-test diagnosis.

2026-07-082

agent-canon-update

iwashita-nozomu/agent-canon

Use when updating AgentCanon itself, refreshing a vendored vendor/agent-canon submodule pin, repairing AgentCanon root runtime views, applying AgentCanon update TODOs, or routing local AgentCanon source commits through a proper AgentCanon branch and PR before parent pin updates.

2026-07-072

agent-eval-accumulation

iwashita-nozomu/agent-canon

Use when accumulated AgentCanon eval evidence is missing, stale, or failing; runs registered eval producers, validates eval family accumulation, and stores evidence through the log archive instead of hand-writing reports.

2026-07-072

name	html-experiment-report
description	Use when producing a browser-readable HTML experiment or Eval report; first decide the primary figure, then plan and run an evidence-backed report renderer while keeping domain authority in the original tool.

HTML Experiment Report

Tool Commands

Use the command packet before applying this skill's workflow:

python3 tools/agent_tools/skill_tool_commands.py show --skill html-experiment-report --format text

Execute the required and task-matching conditional commands that the packet prints.

Read agents/skills/html-experiment-report.md.
Use $structure-planning before renderer or experiment implementation to fix the first figure, source-to-structure map, metric contract, section order, and invalid interpretations.
Use $html-output for polished HTML layout, optional $imagegen assets, existing preview-server reuse, python3 -m http.server --bind 0.0.0.0 publication, curl -fsS validation, and local/external browser URLs.
Survey existing assets first: relevant skills, tool catalog entries, workflow docs, report helpers, previous run artifacts, and existing experiment scripts.
Write a responsibility analysis before implementation: raw evidence owner, renderer owner, domain decision owner, generated artifact path, and the invalid authority drift to avoid.
Name the first figure through the structure contract before running the experiment. Record its question, required data, metric denominator, directionality, and invalid interpretations.
Derive the experiment plan from that figure: evidence-producing command, renderer command, HTML output path, blocked-provider behavior, and validation gates.
Reuse existing producers and helpers. Add a report-specific renderer or adapter only when the HTML artifact needs one.
Keep generated SQLite, JSON, and HTML artifacts under ignored run paths such as reports/agents/<run-id>/; do not make generated HTML a policy truth.
Put the primary figure first in the HTML, followed by source packet, observations, interpretation, limitations, provenance, and next action.
For semantic-index provider comparison, use the figure Provider Delta To Shared Candidate Logic and state that LLM latent vectors may change retrieval/ranking deltas but do not create labels, ownership, or merge authority.
Validate with targeted renderer tests, docs checks for changed Markdown, and catalog/dependency checks for changed tool or skill wiring.
Record closeout tokens: html_experiment_report=complete, html_primary_figure=<figure-name>, structure_contract=<path-or-inline>, html_source_artifact=<path>, html_report_artifact=<path>, html_domain_authority=<tool-or-doc>, and html_output=complete, html_server_mode=<reuse|started|not_requested|blocked>, and html_invalid_interpretations_recorded=yes.