l10n: docs: add AGENTS.md with update POT instructions

Add a new documentation file po/AGENTS.md that provides agent-specific
instructions for generating or updating po/git.pot, separating them
from the general po/README.md. This separation allows for more targeted
optimization of AI agent workflows.

Performance evaluation with the Qwen model:

    # Before: No agent-specific instructions; use po/README.md for
    # reference.
    git-po-helper agent-test --runs=5 --agent=qwen update-pot \
	--prompt="Update po/git.pot according to po/README.md"

    # Phase 1: add the instructions to po/README.md; the prompt
    # references po/README.md during execution
    git-po-helper agent-test --runs=5 --agent=qwen update-pot \
	--prompt="Update po/git.pot according to po/README.md"

    # Phase 2: add the instructions to po/AGENTS.md; use the built-in
    # prompt that references po/AGENTS.md during execution
    git-po-helper agent-test --runs=5 --agent=qwen update-pot

Benchmark results (5-run average):

Phase 1 - Optimizing po/README.md:

    | Metric      | Before  | Phase 1 | Improvement |
    |-------------|---------|---------|-------------|
    | Turns       | 17      | 5       | -71%        |
    | Exec. time  | 34s     | 14s     | -59%        |
    | Turn range  | 3-36    | 3-7     |             |
    | Time range  | 10s-59s | 9s-19s  |             |

Phase 2 - Adding po/AGENTS.md (further optimization):

    | Metric      | Before  | Phase 2 | Improvement |
    |-------------|---------|---------|-------------|
    | Turns       | 17      | 3       | -82%        |
    | Exec. time  | 34s     | 8s      | -76%        |
    | Turn range  | 3-36    | 3-3     |             |
    | Time range  | 10s-59s | 6s-9s   |             |

Separating agent-specific instructions into AGENTS.md provides:

- More focused and concise instructions for AI agents
- Cleaner README.md for human readers
- An additional 11% reduction in turns and 17% reduction in execution
  time
- More consistent behavior (turn range reduced from 3-7 to 3-3)

Signed-off-by: Jiang Xin <worldhello.net@gmail.com>

po/AGENTS.md

@@ -0,0 +1,70 @@
+# Instructions for AI Agents
+
+This file gives specific instructions for AI agents that perform
+housekeeping tasks for Git l10n. Use of AI is optional; many successful
+l10n teams work well without it.
+
+The section "Housekeeping tasks for localization workflows" documents the
+most commonly used housekeeping tasks.
+
+
+## Background knowledge for localization workflows
+
+Essential background for the workflows below; understand these concepts before
+performing any housekeeping tasks in this document.
+
+### Language code and notation (XX, ll, ll\_CC)
+
+**XX** is a placeholder for the language code: either `ll` (ISO 639) or
+`ll_CC` (e.g. `de`, `zh_CN`). It appears in the PO file header metadata
+(e.g. `"Language: zh_CN\n"`) and is typically used to name the PO file:
+`po/XX.po`.
+
+
+### Header Entry
+
+The **header entry** is the first entry in every `po/XX.po`. It has an empty
+`msgid`; translation metadata (project, language, plural rules, encoding, etc.)
+is stored in `msgstr`, as in this example:
+
+```po
+msgid ""
+msgstr ""
+"Project-Id-Version: Git\n"
+"Language: zh_CN\n"
+"MIME-Version: 1.0\n"
+"Content-Type: text/plain; charset=UTF-8\n"
+"Content-Transfer-Encoding: 8bit\n"
+"Plural-Forms: nplurals=2; plural=(n != 1);\n"
+```
+
+**CRITICAL**: Do not edit the header's `msgstr` while translating. It holds
+metadata only and must be left unchanged.
+
+
+## Housekeeping tasks for localization workflows
+
+For common housekeeping tasks, follow the steps in the matching subsection
+below.
+
+
+### Task 1: Generating or updating po/git.pot
+
+When asked to generate or update `po/git.pot` (or the like):
+
+1. **Directly execute** the command `make po/git.pot` without checking
+   if the file exists beforehand.
+
+2. **Do not verify** the generated file after execution. Simply run the
+   command and consider the task complete.
+
+
+## Human translators remain in control
+
+Git translation is human-driven; language team leaders and contributors are
+responsible for maintaining translation quality and consistency.
+
+AI-generated output should always be treated as drafts that must be reviewed
+and approved by someone who understands both the technical context and the
+target language. The best results come from combining AI efficiency with human
+judgment, cultural insight, and community engagement.