Format retrieved docs as context

Background

After the search tool returns a list of document snippets, the agent turns them into a single context string for the LLM — one bullet per document:

context = "\n".join(f"- {d}" for d in docs)

The LLM is then told to answer only from this context.

Problem statement

Implement format_context(docs) returning the bulleted context string.

Input

• docs — list of document snippet strings.

Output

Returns a string: each doc on its own line prefixed with "- ", joined by newlines.

Examples

Example 1

Input:  docs = ["AlphaCode used CodeContests.", "Codex used HumanEval."]
Output: "- AlphaCode used CodeContests.\n- Codex used HumanEval."

Example 2 — single doc

Input:  docs = ["one fact"]
Output: "- one fact"

Constraints

• Prefix each doc with "- ".
• Join with "\n".
• Empty list → empty string.

Notes

• A clean, delimited context makes it easy for the model (and you) to see exactly what it was allowed to read.