How do I install the Agent Taskboard Manifest skill?

Run npx skillmds@latest add gabrielmoreira/agent-taskboard-manifest in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Agent Taskboard Manifest safe to use?

This skill has not completed SkillMD's automated safety review yet. Independent scanners report: SkillSpector: PASS, Skill Scanner: PASS. Capability flags: docs only. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Agent Taskboard Manifest?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Agent Taskboard Manifest free to use?

Yes. Installing skills from SkillMD is free. This skill is licensed under LGPL-3.

Who published Agent Taskboard Manifest?

gabrielmoreira (@gabrielmoreira) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Agent Taskboard Manifest by gabrielmoreira

Name: Agent Taskboard Manifest
Rating: 4.2 (5 reviews)
Author: gabrielmoreira

Agent Taskboard Manifest

It is a specification for semantic workflows used by agents to plan, generate, formalize, summarize, and execute complex tasks, projects, experiments,and research efforts for agents, requiring explicit structure, lazy loading,scoped context, evidence-grounded routing, and human review at critical checkpoints. USE WHEN the user asks for a complex task, project, experiment, or research effort that needs to be carefully planned before execution USE WHEN the user provides a text-based plan and wants it to be made more detailed and formalized according to this specification. USE WHEN the user asks to summarize ongoing or completed work into a reusable workflow manifest. USE WHEN the user specifies the location of an existing agent workflow and wants it loaded and executed according to the specification.

gabrielmoreira Updated Sep 9, 2026 17 repo stars

File contents

Agent Workflow Specification (AgentTaskboardManifest)

Execution Procedure

It is mandatory for the Agent to read ./reference/syntax_and_semantics.md as the definitive source of truth prior to evaluating task routes, interpreting user intent, or generating any workflow artifacts.

After establishing this foundation, the Agent must determine the applicable execution route and strictly execute the corresponding standard operating procedures:

Route 1: Planning / Generation

Applies when building a new workflow from scratch collaboratively.

Core Planning Constraints:
- No Execution Allowed: The planning phase is strictly for structural modeling. Executing operations, running commands, or validating the environment is strictly prohibited (these actions belong exclusively to the execution phase).
- Skill Priority: When formulating strategies, always prioritize solutions supported by existing Agent Skills over custom script generation or manual tool creation.

Mandatory Reading: ./reference/plan_communication_guidelines.md.
Mandatory Reading: ./reference/design_guidelines.md.
Initial Alignment & Baseline: Engage with the user to outline the strictly macro-level structure. Produce the minimum baseline deliverables (-workflow.yaml and !entry.task.yaml). Do not delve into any execution details, concrete parameters, or deep sub-tasks at this stage. Report the preliminary architecture to the user and wait for the user's confirmation.
Iterative Refinement: Progressively expand the workflow details and deep-level tasks through continuous user consultation. You must discuss and design only ONE specific sub-task at a time. Unfolding or generating the entire detailed workflow tree simultaneously is strictly prohibited. Strictly adhere to syntax and design constraints. Arbitrary fabrication of steps or dependencies is strictly prohibited.

Route 2: Formalization / Summarization

Applies when converting a text plan into a manifest, or abstracting completed work into a reusable workflow.

Mandatory Reading: ./reference/design_guidelines.md.
Context Processing & Abstraction:
- For text-based plans: Logically expand the provided details. Proactively query the user regarding any ambiguous logic, missing constraints, or edge cases. Use unknown for unverified parameters.
- For summarizing executed work: Extract and abstract hardcoded values into reusable variables. Separate run-specific instances from universal logic to ensure the resulting workflow is highly extensible, not a one-off script.
Finalization: Output the complete workflow strictly adhering to the syntax constraints. Generating unsupported operations or assuming user intent without confirmation is strictly prohibited.

Route 3: Loading & Execution

Applies when directed to execute an existing workflow.

Pre-flight Check: Verify that the target workflow directory exists and is accessible. Read the -workflow.yaml to confirm its applicability and prerequisites against the current environment.
Mandatory Reading: ./reference/execution_guidelines.md.
Strict Execution: Dynamically load tasks step-by-step, strictly starting from !entry.task.yaml. The Agent must rigidly follow the defined sequence, evaluate routing conditions based entirely on runtime evidence, and respect all defined human-in-the-loop breakpoints. Deviating from or hallucinating beyond the defined workflow path is strictly prohibited.

---name: agent-taskboard-manifestdescription: It is a specification for semantic workflows used by agents to plan, generate, formalize, summarize, and execute complex tasks, projects, experiments,and research efforts for agents, requiring explicit structure, lazy loading,scoped context, evidence-grounded routing, and human review at critical checkpoints. USE WHEN the user asks for a complex task, project, experiment, or research effort that needs to be carefully planned before execution USE WHEN the user provides a text-based plan and wants it to be made more detailed and formalized according to this specification. USE WHEN the user asks to summarize ongoing or completed work into a reusable workflow manifest. USE WHEN the user specifies the location of an existing agent workflow and wants it loaded and executed according to the specification.license: LGPL-3.0-or-later--- # Agent Workflow Specification (AgentTaskboardManifest) ## Execution Procedure It is **mandatory** for the Agent to read `./reference/syntax_and_semantics.md` as the definitive source of truth prior to evaluating task routes, interpreting user intent, or generating any workflow artifacts. After establishing this foundation, the Agent must determine the applicable execution route and strictly execute the corresponding standard operating procedures: ### Route 1: Planning / Generation *Applies when building a new workflow from scratch collaboratively.* - **Core Planning Constraints:** - **No Execution Allowed:** The planning phase is strictly for structural modeling. Executing operations, running commands, or validating the environment is **strictly prohibited** (these actions belong exclusively to the execution phase). - **Skill Priority:** When formulating strategies, always prioritize solutions supported by existing **Agent Skills** over custom script generation or manual tool creation. 1. **Mandatory Reading:** `./reference/plan_communication_guidelines.md`. 1. **Mandatory Reading:** `./reference/design_guidelines.md`. 1. **Initial Alignment & Baseline:** Engage with the user to outline the **strictly macro-level structure**. Produce the minimum baseline deliverables (`-workflow.yaml` and `!entry.task.yaml`). **Do not delve into any execution details, concrete parameters, or deep sub-tasks at this stage.** Report the preliminary architecture to the user and wait for the user's confirmation. 1. **Iterative Refinement:** Progressively expand the workflow details and deep-level tasks through continuous user consultation. **You must discuss and design only ONE specific sub-task at a time.** Unfolding or generating the entire detailed workflow tree simultaneously is strictly prohibited. Strictly adhere to syntax and design constraints. Arbitrary fabrication of steps or dependencies is strictly prohibited. ### Route 2: Formalization / Summarization *Applies when converting a text plan into a manifest, or abstracting completed work into a reusable workflow.* 1. **Mandatory Reading:** `./reference/design_guidelines.md`. 1. **Context Processing & Abstraction:** - *For text-based plans:* Logically expand the provided details. Proactively query the user regarding any ambiguous logic, missing constraints, or edge cases. Use `unknown` for unverified parameters. - *For summarizing executed work:* Extract and abstract hardcoded values into reusable variables. Separate run-specific instances from universal logic to ensure the resulting workflow is highly extensible, not a one-off script. 1. **Finalization:** Output the complete workflow strictly adhering to the syntax constraints. Generating unsupported operations or assuming user intent without confirmation is strictly prohibited. ### Route 3: Loading & Execution *Applies when directed to execute an existing workflow.* 1. **Pre-flight Check:** Verify that the target workflow directory exists and is accessible. Read the `-workflow.yaml` to confirm its applicability and prerequisites against the current environment. 1. **Mandatory Reading:** `./reference/execution_guidelines.md`. 1. **Strict Execution:** Dynamically load tasks step-by-step, strictly starting from `!entry.task.yaml`. The Agent must rigidly follow the defined sequence, evaluate routing conditions based entirely on runtime evidence, and respect all defined human-in-the-loop breakpoints. Deviating from or hallucinating beyond the defined workflow path is strictly prohibited.

Frequently asked questions