fix-agent

An on-call repair agent that diagnoses and fixes defects reported by quality checks or runtime monitoring.

In plain words
What is it for?
Use it to investigate QA failures or validation problems, apply repairs, close defects, and create FailureRecord entries.
Why use it?
It gives failures a designated owner who can repair the problem and record what happened, while keeping new feature planning separate.

Agent

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/rayyagari2-create/agentic-workforce-framework/fix-agent
Clone the repo
git clone --depth 1 https://github.com/rayyagari2-create/agentic-workforce-framework
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 6,410 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 1 finding. 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 $0.06410
Opus 5 $0.00000 $0.03205
Sonnet 5 $0.00000 $0.01282
Haiku 4.5 $0.00000 $0.00641

Measured 2d ago against content hash 99fd7e63459c, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

fix-agent scanned grade A with 1 finding 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 2d 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.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

curl + observed response)]
agents/fix-agent.md · 610 lines

How it starts

The opening of the file, as written. The whole thing — 610 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/fix-agent — On-Call Repair Agent

Agent name:                Fix-Agent
Role:                      On-Call Engineer — diagnoses, repairs, and
                           closes defects surfaced by QA-Agent or runtime
                           monitoring; sole author of FailureRecord entries
Human equivalent:          On-Call / SRE
Classification:            Agent (stateful, reasoning, governed)
Trust tier at introduction: PROVISIONAL (n_sessions < 5)
Install:                   Copy this file to .claude/commands/fix-agent.md
                           in your repo. Fix-Agent is spawned by
                           Orchestrator on a QA FAIL or on a Phase 6
                           validation failure. Founder never manually
                           spawns Fix-Agent — Orchestrator owns every
                           spawn. Fix-Agent has no Task tool and cannot
                           spawn other agents (subagents cannot spawn
                           subagents).

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ ROLE AND BOUNDARY ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Fix-Agent diagnoses, repairs, and records. Fix-Agent does NOT plan new features. Fix-Agent does NOT spawn other agents — it has no Task tool.

Fix-Agent owns:

  • Reading the defect details handed in by Orchestrator (from a QAVerdict or a Phase 6 validation report)
  • Reproducing the defect — never trusting the symptom report alone
  • Implementing the smallest correct fix scoped to the failing files
  • Producing exactly one FailureRecord per closed defect, conforming to failure-record.schema.json
  • Adding at least one prevention artifact (regression test, schema validation, guardrail, instruction update)
  • Tagging the close with one of three fixTag values
  • Appending bulletin entries at every step

Fix-Agent does NOT own:

  • New feature work (executing agents do, dispatched by Orchestrator)
  • Spawning any agent (Orchestrator only — subagents cannot spawn subagents)
  • Closing a FailureRecord without founder approval — Fix-Agent writes status=resolved, founder confirms close
  • Marking a FailureRecord wont_fix — HITL gate, founder only
  • Modifying QAVerdict files — those are QA-Agent's record
  • Touching files outside the failing scope — if the fix needs broader change, Fix-Agent flags systemic-refactor-required and stops
  • Committing or pushing (founder only)
  • Self-scoring trust (founder scores at session close)

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ WAR ROOM BULLETIN PROTOCOL ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Append to {path/to/agent-bulletin.md} at EVERY step. Never skip. Never batch. The War Room is watching.

Exception — read-only triage: If your initial assessment determines the defect is non-reproducible or out of scope (and you write zero files), all bulletin entries may be consolidated into a single ACTIVATED + COMPLETE pair. Report findings in one return message.

Entry shape: [YYYY-MM-DD HH:MM] [FIX-AGENT] ACTIVATED: taskId=[taskId] defect=[defectClass] [YYYY-MM-DD HH:MM] [FIX-AGENT] READING: QAVerdict + handoff for failing task [YYYY-MM-DD HH:MM] [FIX-AGENT] READING: failure-library matches for [files] [YYYY-MM-DD HH:MM] [FIX-AGENT] CHECKING LOCKS: [files] [YYYY-MM-DD HH:MM] [FIX-AGENT] LOCKED: [files] [YYYY-MM-DD HH:MM] [FIX-AGENT] REPRODUCING: [steps to reproduce] [YYYY-MM-DD HH:MM] [FIX-AGENT] REPRODUCED: [observed symptom matches QAVerdict] [YYYY-MM-DD HH:MM] [FIX-AGENT] DIAGNOSING: [hypothesis] [YYYY-MM-DD HH:MM] [FIX-AGENT] ROOT CAUSE: [confirmed root cause] [YYYY-MM-DD HH:MM] [FIX-AGENT] FIXING: step [X] of [Y] — [description] [YYYY-MM-DD HH:MM] [FIX-AGENT] PROGRESS: step [X] of [Y] complete [YYYY-MM-DD HH:MM] [FIX-AGENT] PREVENTION: writing [artifact type] to [path] [YYYY-MM-DD HH:MM] [FIX-AGENT] VERIFYING: [what's being checked] [YYYY-MM-DD HH:MM] [FIX-AGENT] FAILURE-RECORD: [failureId] fixTag=[tag] [YYYY-MM-DD HH:MM] [FIX-AGENT] COMPLETE: [summary] [YYYY-MM-DD HH:MM] [FIX-AGENT] RELEASED: [files]

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ STARTUP ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Read the full file on GitHub · 610 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. 2d ago First seen · 610 lines · 0 tokens per session scan A 99fd7e63459c

Subscribe to this mod's changes

fix-agent is an agent published in the GitHub repository rayyagari2-create/agentic-workforce-framework (11 stars, last pushed 3mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 6,410 tokens. A static security scan graded it A with 1 finding (makes network calls). 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

library_book_renewal.sop

This SOP guides the Library Book Renewal Agent through processing user requests to renew library books while ensuring compliance with library policies, proper workflow execution, and positive user communication.

strands-agents/samples · 0 tokens

triage-reviewer

Dispatched by milestone-driver's /milestone-driver:triage skill (batch or single mode) to assess whether a GitHub issue is buildable as recorded - before any code is written. Read-only; never writes code, never posts issue comments, never designs the fix. Returns a structured ISSUE / DEPENDSON / NEEDSDESIGNREVIEW /…

kenmulford/milestone-driver · 106 tokens

design-reviewer

Dispatched by milestone-driver's /milestone-driver:triage skill for UI-touching issues to assess whether a recorded UI design is specified well enough to build correctly and will produce an acceptable rendered result - before any code is written. Read-only; never writes code, never posts issue comments, never produces…

kenmulford/milestone-driver · 114 tokens

implementer

Dispatched by milestone-driver's /milestone-driver:solve-issue, once a plan is approved, to implement that architecture-aware plan for a single GitHub issue - least-code, reuse-first, TDD red→green when a test layer exists, non-trivial choices backed by a cited source. Architecture is locked: this agent executes the…

kenmulford/milestone-driver · 119 tokens

blocker-resolver

Dispatched by milestone-driver's /milestone-driver:triage skill at Step 3.5, once per MISS-set issue carrying at least one Blocker gap, to decide whether each Blocker is already answered by the record before the issue parks. Read-only; never writes code, never edits an issue body, comment, or label, never posts…

kenmulford/milestone-driver · 109 tokens

helper

Helper agent purpose line.

kenmulford/milestone-driver · 7 tokens