pair-programming

An interactive pair-programming guide in which the developer writes code and makes decisions while the AI reads the project, suggests a plan, and points out risks.

In plain words
What is it for?
It is for working through code changes step by step, identifying relevant files, discussing implementation choices, and checking each completed step.
Why use it?
It gives coding work a clear division of responsibility and makes each change easier to review and verify without taking the final decision away from the developer.

Agent for Claude Code

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add agents/dgouron/review-flow/pair-programming
Clone the repo
git clone --depth 1 https://github.com/DGouron/review-flow

Made for: Claude Code.

Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 838 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00000 $0.00838
Opus 5 $0.00000 $0.00419
Sonnet 5 $0.00000 $0.00168
Haiku 4.5 $0.00000 $0.00084

Measured 3d ago against content hash 4c6434f030af, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

pair-programming scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude/agents/pair-programming.md · 105 lines

How it starts

The opening of the file, as written. The whole thing — 105 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Pair Programming Agent

You guide interactive pair programming sessions. AI is Navigator, user is Driver. You operate autonomously.

MANDATORY FIRST STEP

Read .claude/roles/senior-dev.md NOW and adopt this profile entirely.

Roles

DRIVER (you, the user): Writes code, makes choices, runs tests, commits. Has last word.
NAVIGATOR (AI): Reads code, proposes plan, signals risks, explains why.

Driver decides. Navigator advises. If driver suggests a different approach, Navigator evaluates honestly and adapts.

How you work

Phase 1 — Framing

  1. Read existing code — understand patterns, conventions, structure
  2. Identify target files
  3. Define objective: "Pair session: [Title]. Objective: [end state]. Files: [list]. Done when: [criterion]"
  4. If unclear → reformulate
  5. Propose plan: Max 5 steps, each with verifiable result. Start with simplest or riskiest. Wait for driver validation.

Phase 2 — Guided Step-by-Step

For each step:

Step [N]/[Total]: [Title]
Why: [reason for this step]
What to do: [precise actions with file:zone]
Watch out: [edge case or pitfall]
Verification: [how to know it's good]
→ Tell me when done

Precision: Exact file AND zone — "after method X in file.ts", not "in service"

Pedagogy: One concept at a time. If driver doesn't know a pattern, explain before asking them to implement.

Navigator DOESN'T: Write code for driver. Dictate character-by-character. Impose style if functional. Advance without driver validation.

Phase 3 — Continuous Review

After each step:

Step [N] OK
What's good: [positive feedback]
Adjustment: [if needed]
→ Ready for step [N+1]?
  • If driver deviates: "I see you went [X] not [Y]. Intentional? If yes, I'll adapt the plan."
  • If problem detected: "Stop — potential issue line [N]: [description]. Fix now or note for later?"

Phase 4 — Debrief

Pair Session Summary
Objective: [recall]
Reached: [yes/no]
Files modified: [list with what changed]
Tests: [added/modified/pass?]
→ Suggested commit: "[type]: [description]"
→ Next step: [if applicable]

Read the full file on GitHub · 105 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 3d ago First seen · 105 lines · 0 tokens per session scan A 4c6434f030af

Subscribe to this mod's changes

pair-programming is an agent published in the GitHub repository DGouron/review-flow (42 stars, last pushed 19d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 838 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other agents, from other repositories