Agent Mentor Learn

Sources

The key claims in this course rest on the sources below. The lessons cite them with [^Sn]; the exact excerpts sit under each entry.

S1 — How we built our multi-agent research system — Anthropic Engineering

URL: https://www.anthropic.com/engineering/multi-agent-research-system

  • authority: authoritative-guide

Anthropic's engineering retrospective on its multi-agent research system, and this course's core basis: agents are stateful and errors compound; you can't restart from scratch — you have to resume from where the error hit; reliability comes from pairing the model's adaptability with deterministic safeguards like retry logic and regular checkpoints; and at deploy time a running agent may be anywhere in its process.

Key quote:

"Agents can run for long periods of time, maintaining state across many tool calls." "Agents are stateful and errors compound." "Without effective mitigations, minor system failures can be catastrophic for agents." "When errors occur, we can't just restart from the beginning: restarts are expensive and frustrating for users." "Instead, we built systems that can resume from where the agent was when the errors occurred." "This means we need to durably execute code and handle errors along the way." "the adaptability of AI agents built on Claude with deterministic safeguards like retry logic and regular checkpoints" "letting the agent know when a tool is failing and letting it adapt works surprisingly well" "we use rainbow deployments to avoid disrupting running agents, by gradually shifting traffic from old to new versions" "whenever we deploy updates, agents might be anywhere in their process." "agents summarize completed work phases and store essential information in external memory before proceeding to new tasks"

S2 — Checkpointing — Claude Code Docs

URL: https://code.claude.com/docs/en/checkpointing

  • authority: official-docs

The official docs for Claude Code's checkpointing, showing the complete product-grade shape of checkpoint/rewind/fork: an automatic checkpoint before every user prompt, saved with the conversation, conversation and code restorable separately, bash-made changes untracked, and no substitute for version control. This course uses it as a product-grade reference point, not as the tool being taught.

Key quote:

"checkpointing automatically captures the state of your code before each user prompt." "Every user prompt creates a new checkpoint" "Claude Code saves checkpoints with the conversation, so you can still run /rewind after you resume a session" "Restore code and conversation: revert both code and conversation to that point" "Restore conversation: rewind to that message while keeping current code" "Restore code: revert file changes while keeping the conversation" "Exploring alternatives: try different implementation approaches without losing your starting point" "Recovering from mistakes: quickly undo changes that introduced bugs or broke functionality" "Checkpointing does not track files modified by bash commands." "Only direct file edits made through Claude's file editing tools are tracked." "Checkpoints are designed for quick, session-level recovery." "continue using version control, such as Git, for commits, branches, and long-term history." "To branch off and try a different approach while preserving the original session intact, use /branch or claude --continue --fork-session"

S3 — Building Effective AI Agents — Anthropic Engineering

URL: https://www.anthropic.com/engineering/building-effective-agents

  • authority: authoritative-guide

Anthropic's authoritative article on agent patterns. This course uses it for: agents pausing at checkpoints for human feedback, the higher costs and compounding errors that autonomy brings, the limited trust across many turns, and the proportion rule — add complexity only when it demonstrably improves outcomes.

Key quote:

"Agents can then pause for human feedback at checkpoints or when encountering blockers." "The autonomous nature of agents means higher costs, and the potential for compounding errors." "The LLM will potentially operate for many turns, and you must have some level of trust in its decision-making." "it's also common to include stopping conditions (such as a maximum number of iterations) to maintain control." "you should consider adding complexity only when it demonstrably improves outcomes."

S4 — Handle tool calls — Claude API

URL: https://platform.claude.com/docs/en/agents-and-tools/tool-use/handle-tool-calls

  • authority: official-docs

Gives the field spec and pairing rule for the tool_use and tool_result content blocks: every tool_use must get a matching tool_result, all returned together. The resume lesson depends on this rule to handle the dangling call left when a crash lands between tool execution and ledger write.

Key quote:

"id: A unique identifier for this particular tool use block. ... name: The name of the tool being used. input: An object containing the input being passed to the tool, conforming to the tool's input_schema. ... tool_use_id: The id of the tool use request this is a result for. ... is_error (optional): Set to true if the tool execution resulted in an error." "Whichever strategy you use, return one tool_result for each tool_use block, all together in the next user message. Match each result to its call with tool_use_id, and put every tool_result block before any text content in that message."

S5 — The 2026 Agent Engineering Roadmap — GitHub (codejunkie99/agent-roadmap-2026)

URL: https://github.com/codejunkie99/agent-roadmap-2026

  • authority: blog

A community open-source roadmap organized around harness engineering (AI-assisted writing). This course takes only its framing claims: persistence is one component of the harness, and its job is "checkpoint state every node so you can resume, rewind, fork." Note: its percentage thresholds, token figures, benchmark scores, salary numbers, and other specifics are never cited as fact.

Key quote:

"the harness is the union of:" "persistence. Checkpoint state every node so you can resume, rewind, fork." "Same model, different harness, completely different result."