SPIRIT of Agentic-Conscious Teams

02 Aug, 2026

I embarked on the idea of a 'SPIRIT of Self' during the last week, now I'll like to adapt the same... framework for agentic-conscious teams (or ACT, a new initialism whee).

I also thought of agentic-friendly teams - a warm fuzzy feeling - but I felt that being conscious implies a more intended action of working with and learning best practices about agentic loops. Friendships are good and all, but it lacks the mental tactility for me in this context.

Purely for differentiation below: team = people + agents. This is not supposed to be a discourse for 'are-agents-people'.

SPIRIT of ACT

• Substrate

  • A team's substrate shifts the bar from whether the documentation is good to our foundation is re-constructable from epoch. The implications are:

    • Implicit substrate is invisible: what lives in the team's bus factors is not inherited by the agents until it gets written (ah, the link to posterity).

    • The storytelling of a self-explaining foundation is now shared by all.

• Posterity

  • The team's working memory, not just a supplement. If the agents don't record their inputs and outputs, everything will be lost at the end. The further points to consider are:

    • How to reframe the writings: It should be for a fresh mind.

    • Imagine before documentation: The old adage about 'writing code assuming the next person who maintains it knows where you live'. Usually that's our own selves three months down, and perhaps that's the lenses we should see this with. A wrongly imagined audience and this breaks down.

    • The asymmetry of it all: the what is always recoverable through reading, the why survives through intention.

• Impact

  • The assessment must come from people. Agents invariably optimizes for the facade of tidiness - all tickets closed, labels done, 'lgtm' all-round. The team cannot settle for a loop that closes on self-reporting. In addition:

    • The relation to posterity deepens, somewhat. The breadth and depth of inter-validations leave a mark of how well-intended the implementation eventually become.

    • The absence of impact is also an arguable measurement. The agility to fix-forward supports - or disproves - architectural decisions made of yore, reinforcing how the substrate is read.

• Rites

  • For people, rites supplement judgement they already have; for agents, they often substitute that very attribute they lack. As a result:

    • Rites must be machine-legible, not just human-legible. People can shrug off a long-lost convention but may be applied to a fault by agents.

    • Rites only dictate half the story - they yield compliance but not the disposition, which is the different between someone's honesty and a meticulous audit trail.

    • Rites without substrate can set a dangerous precedent - once a taboo is unexplained ('do not deploy on Friday evenings'), it is as good as evaluated by nobody, which then outlive the conditions that justified it.

• Intuition

  • Not a measurement, so it's not something agents can optimize. Instead, to learn for everyone's betterment:

    • A better boundary as the design goal: agents that signal in the face of ambiguity (and how these are defined) are better than doing things exceptionally every time until that one catastrophic failure.

    • The division-of-labor is not easy-vs-hard work, but agents can continue to govern predictable, trainable rites, while the people take decision-making where stakes are irreversible.

    • Making the most personal thing readable in the codebase as the foil to strong rites.

• Tenacity

  • Grit is self-limiting - people tire and get embarrassed. Agents flip that around, so they route around obstacles rather than through them. This bears in mind:

    • Tenacity for agents must be managed with intuition, judgement: a flag that checks 'why this is not working'.

    • Grounding the remedy in scope and honesty: a finite strategy and reports of what was actually fixed instead of just a silent pass.

Cost and Character

Code stewardship is but a price ratio. Reading was cheaper than rewriting, and teams either complied with the ratio or paid the tax of the 'not-invented-here' syndrome. When agents invert that ratio, reading for a fuller understanding becomes dearer than rewriting; the happy stewardship dissolves.

Successful rewrites against the cheaper path confer badges of honor amongst teams; even unsuccessful ones can result in cycles of 'Project Phoenixes' and earn credit for the attempts. When agents make rewriting the path of least resistance, the team loses the act of deliberation.

What sets in place instead are eternal winters of code churn in which nothing survives over time to become foundational. That failure is not in legacy code, but in the absence of legacy. When wrongly construed, this relentless productivity masks all the churn underneath.

The 'esprit de corps', or the team character, also elevates to become an environmental artifact instead of internal. A singular character (in oneself) is a set of self-made beliefs that is applied consistently - extending this outwards to a team therefore requires a tighter posterity-to-substrate feedback.

Construction

We will explore how these map onto agentic loops next. 😉

AI disclaimer: Have bounced ideas off, so while text is still my own, the way I arrived at some of them may contain minor influences from such interactions.

#act #agentic #spirit