Skip to content

Commit 2b8daae

Browse files
ATXcoderATXcoder
committed
future post updates
1 parent 03d384e commit 2b8daae

File tree

4 files changed

+196
-3
lines changed

4 files changed

+196
-3
lines changed

‎_drafts/my-five-time-saving-automations.md‎

Lines changed: 6 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -4,3 +4,9 @@ description: "Five simple, real-world automations I personally use at work and a
44
categories: [Automation, Productivity]
55
tags: [automation, power-automate, n8n, node-red, productivity, workflows, it admin]
66
---
7+
8+
## Outlook to Todoist
9+
10+
## Automatic Outlook Out Of Office
11+
12+
## Nightly Borgbase Backup

‎_drafts/spotting-automation-technical-debt.md‎

Lines changed: 187 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,193 @@
11
---
22
title: "How to Spot Automation That Will Become Technical Debt"
33
description: "Learn the warning signs that an automation is quietly becoming technical debt, why it happens, and how to fix it before it turns into a long-term maintenance problem."
4+
image: /assets/img/techincal-debt.jpg
45
categories: [Automation, Architecture]
56
tags: [automation, technical-debt, power-automate, system-design, it-architecture, maintainability]
67
---
8+
9+
**Just because it runs doesn’t mean it’s healthy**
10+
11+
Automation should make work *easier*, not slower. Too often we build workflows, bots, and flows that technically still run — but quietly become obstacles. They work… but they hold you back.
12+
13+
That’s what I mean by **automation debt**:
14+
> Automation that still runs but actively slows you down.
15+
16+
Everyone builds this stuff. I’ve built it. You’ve built it. Entire teams have unknowingly let it accumulate.
17+
18+
In this post, I’ll walk through the common signs of automation debt — and what you can do about it without rewriting everything.
19+
20+
---
21+
22+
## What Is Automation Debt?
23+
24+
Automation debt isn’t about broken flows or errors. It’s about *friction*.
25+
26+
It’s automation that:
27+
- still executes,
28+
- still delivers value,
29+
- but takes more effort to maintain than it should.
30+
31+
It creeps in slowly, one “quick fix” at a time.
32+
33+
Let’s look at the signals.
34+
35+
---
36+
37+
## 1. No One Knows Why It Exists
38+
39+
### Signs to Watch For
40+
- There’s no description or documentation.
41+
- Nobody owns it.
42+
- Names are vague — think `Flow 21`, `Copy of Approval`.
43+
44+
You wind up with automations that feel like relics. They *work*, but nobody dares change them.
45+
46+
**If you can’t explain why an automation exists, it’s already debt.**
47+
48+
---
49+
50+
## 2. It Depends on a Single Person’s Account
51+
52+
Maybe it runs under your credentials. Maybe it uses someone’s token.
53+
54+
### Why This Is Risky
55+
- People rotate passwords.
56+
- Folks change roles or leave.
57+
- MFA and policy changes can silently break credentials.
58+
59+
When that account goes stale, your automation often doesn’t crash loudly — it just stops doing what you expect.
60+
61+
**Automation shouldn’t hinge on one human being.**
62+
63+
---
64+
65+
## 3. Hardcoded Values Everywhere
66+
67+
> “Just change this one thing…”
68+
> Famous last words.
69+
70+
### What This Looks Like
71+
- IDs and GUIDs baked into actions
72+
- Environment-specific URLs
73+
- Paths that only work in one context
74+
75+
Every time something changes, you repeat the same manual hunt-and-replace.
76+
77+
**Configuration should be external, not hardcoded.**
78+
79+
---
80+
81+
## 4. No Failure Visibility
82+
83+
If automation fails and nobody notices until a user complains… you lack monitoring.
84+
85+
### Common Symptoms
86+
- No logs
87+
- No alerts
88+
- No success/failure tracking
89+
- You only find out when somebody reports it
90+
91+
That means you’re always reacting instead of *seeing problems early*.
92+
93+
**If users are your monitoring system, you don’t have monitoring.**
94+
95+
---
96+
97+
## 5. It Grew Without Design
98+
99+
Automation often starts simple.
100+
101+
Then:
102+
- More steps
103+
- Nested conditions
104+
- Repeated blocks
105+
- No modular pieces
106+
107+
Eventually, it’s a tangled mess.
108+
109+
> Automation doesn’t rot.
110+
> **It spawns.**
111+
112+
At that point, trying to change it can feel like defusing a bomb.
113+
114+
---
115+
116+
## 6. Nobody Knows the Cost
117+
118+
Automation isn’t free — technically or financially.
119+
120+
### Things People Miss
121+
- Licensing usage and costs
122+
- API rate limits
123+
- Performance impact
124+
- What happens at scale
125+
126+
If you can’t answer:
127+
- “What is this costing?”
128+
- “What if usage increases?”
129+
130+
…you’re operating blind.
131+
132+
---
133+
134+
## How to Fix Automation Debt (Without Rewriting)
135+
136+
The good news? Most automation debt can be *reduced* — without a full rewrite.
137+
138+
### Assign an Owner
139+
Every automation should have:
140+
- Someone responsible
141+
- A clear point of contact
142+
143+
Ownership brings clarity.
144+
145+
---
146+
147+
### Externalize Configuration
148+
Move IDs, URLs, and environment specifics out of the logic and into variables, config lists, or parameters.
149+
150+
That makes updates cleaner and safer.
151+
152+
---
153+
154+
### Add Basic Health Reporting
155+
You don’t need a fancy dashboard.
156+
157+
Start with:
158+
- Simple success/failure messages
159+
- Centralized logging
160+
- Critical failure alerts
161+
162+
Even basic visibility is better than none.
163+
164+
---
165+
166+
### Document Just Enough
167+
You don’t need a novel — just the essentials:
168+
169+
- What does this do?
170+
- Who uses it?
171+
- What breaks if it stops?
172+
173+
Future you will thank present you.
174+
175+
---
176+
177+
## Closing Thoughts
178+
179+
Healthy automation is:
180+
181+
- *Boring*
182+
- *Predictable*
183+
- *Visible*
184+
185+
It doesn’t require heroics.
186+
It doesn’t make you nervous to open it.
187+
188+
If you’re afraid to touch a workflow…
189+
**It’s already technical debt.**
190+
191+
---
192+
193+
Want a checklist or a follow-up on **refactoring automation safely**? Just ask.

‎_posts/2025-12-20-automate-out-of-office.md‎

Lines changed: 3 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -13,7 +13,7 @@ I always seemed to forget to turn mine on. While I can enable it using my smartp
1313

1414
## The Flow
1515

16-
This automation is easy to set up and takes just a few minutes. We’ll build a Power Automate flow triggered when it detects a calendar event starting in 15 minutes. It will check if the event is marked as "out of office" and, if so, schedule your Out of Office replies to start and end with the event.
16+
This automation is easy to set up and takes just a few minutes. We’ll build a Power Automate flow triggered when it detects a calendar event starting in 30 minutes. It will check if the event is marked as "out of office" and, if so, schedule your Out of Office replies to start and end with the event.
1717

1818
## Trigger
1919

@@ -33,8 +33,8 @@ Add the [Set up automatic replies](https://learn.microsoft.com/en-us/connectors/
3333

3434
Set the action’s parameters as follows:
3535

36-
- **Status**: Scheduled
37-
- **External Audience**: None (this means replies will only be sent to internal recipients)
36+
- **Status**: "Scheduled" - This way the replies are only sent out between the start and end times. Outlook takes care of turning them off once the end time is reached.
37+
- **External Audience**: "None" - This means replies will only be sent to internal recipients.
3838

3939
Under *Advanced Parameters*, set the following using dynamic content:
4040

‎assets/img/techincal-debt.jpg‎

1.62 MB
Loading

0 commit comments

Comments
 (0)